Unmanned aerial vehicle safety communication method based on IRS

Abstract

Coordination is carried out through an IRS deployed between a communication unmanned aerial vehicle A and a communication unmanned aerial vehicle B, the IRS is carried on an unmanned aerial vehicle C, an eavesdropping unmanned aerial vehicle E exists in a communication system, the phase of a signal sent by the unmanned aerial vehicle A is adjusted, the unmanned aerial vehicle A is aligned with the unmanned aerial vehicle B, the signal forwarding process is completed, and the safety rate of the system is improved; wherein the safety rate R of the communication system is related to w and theta, and the optimal w and theta are sequentially obtained by adopting a method of alternately optimizing w and theta, so that the maximum value of the safety rate R of the system is calculated. The invention aims to provide an IRS-based unmanned aerial vehicle safety communication method aiming at the defects in the prior art, a scheme of combining intelligent reflector beam forming and phase adjustment is planned, and the safety performance of a system physical layer is improved by obtaining an optimal solution of beam forming and phase adjustment.

Description

Technical Field

[0001] The present invention relates to the technical field of virtual synchronous generator control, and in particular to an IRS-based unmanned aerial vehicle (UAV) secure communication method. Background Art

[0002] With the rapid development of electronic attack technology, the external electromagnetic environment of drones is becoming increasingly complex and harsh, the intelligence level of malicious nodes is constantly improving, and the physical layer security issues of drone systems are becoming increasingly prominent. The computational complexity of cracking passwords in traditional authentication and encryption algorithms based on cryptographic systems ensures the effectiveness of the algorithms. However, on the one hand, this security based on computational complexity has not been perfectly mathematically proven; on the other hand, the improvement of computer computing power and computing speed in the future will further challenge the security of public key cryptographic systems.

[0003] At present, physical layer security technology can mainly improve the security performance of communication systems through signal processing technologies such as beamforming and artificial noise and reasonable resource management. However, in drone communications, the nodes move at high speed and are random, the channel state changes rapidly, and wireless resources are increasingly scarce. At the same time, the deep integration of communication countermeasure technology and emerging technologies such as artificial intelligence, the types of attacks are constantly changing, and the level of intelligence is constantly improving. Traditional physical layer security-based methods are difficult to respond flexibly.

[0004] In essence, there are several reasons: first, the physical layer security threats faced by drone communication networks are becoming increasingly severe; second, the mobility of drone communication system nodes exacerbates the randomness of the channel; third, the physical layer security transmission coverage of drones is poor. In summary, the issue of drone security communication has become a problem that needs to be solved urgently.

[0005] In recent years, a wireless technology called Intelligent Reflecting Surface (IRS) has been widely studied, bringing new ideas to solving the physical layer security issues of drone systems. Due to the particularity of drone communication systems, the propagation of signals is essentially random and largely uncontrollable. The intelligent reflecting surface can reconstruct the wireless propagation environment by controlling reflection through software. Specifically, the IRS is a plane composed of a large number of low-cost passive reflective units, each of which can independently change the amplitude or phase of the incident signal, thereby assisting in achieving fine-grained three-dimensional beamforming.

[0006] The use of intelligent reflective surfaces can improve the coverage, transmission rate, and physical layer security performance of drone communication systems. In recent years, drone communication systems assisted by intelligent reflective surfaces have received increasing attention.

[0007] For example, the intelligent reflective surface in document 1 [L. Yang, F. Meng, J. Zhang, et al. On the performance of RIS-assisted dual-hop UAV communication systems [J]. IEEE Transactions on Vehicular Technology, 2020, 69 (9): 10385–10390] is used as a relay-assisted UAV system. From the simulation results, it can be seen that the coverage and reliability of the UAV system deployed with the intelligent reflective surface are significantly improved.

[0008] In reference 2 [S.Li, B.Duo, X.Yuan, et al. Reconfigurable intelligent surface assisted UAV communication: joint trajectory design and passive beamforming [J]. IEEE Wireless Communications Letters, 2020, 9 (5): 716–720], a UAV is used as an aerial mobile base station to provide services to ground users with the assistance of a smart reflector. The average achievable rate is maximized by jointly optimizing the UAV flight trajectory and the passive beamforming of the smart reflector, and a closed-form solution for the phase offset matrix of the smart reflector under any given UAV flight trajectory is given.

[0009] In reference 3 [Q. Zhang, W. Saad, M. Bemmnis. Reflections in the sky: millimeter wave communication with UAV-carried intelligent reflectors [C]. 2019 IEEE Global Communications Conference (GLOBECOM), Waikoloa, HI, USA, 2019, 1-6.], intelligent reflectors are placed on drones to provide relay services to ground users without LoS. Reference 3 proposes a downlink transmission capacity maximization algorithm based on reinforcement learning.

[0010] Reference 4 [L.Ge, P.Dong, H.Zhang, et al. Joint beamforming and trajectory optimization for intelligent reflecting surfaces-assisted UAV communications [J]. IEEE Access, 2020, 8: 78702–78712] designed a scenario with multiple intelligent reflective surfaces and a multi-antenna UAV, and maximized the received power by jointly optimizing the passive beamforming and transmit beamforming of the intelligent reflective surfaces.

[0011] Reference 5 [M.Hua, L.Yang, Q.Wu, et al. UAV-assisted intelligent reflecting surface symbiotic radio system [EB / OL]. https: / / arxiv.org / abs / 2007.14029, January, 2021] studied the UAV-assisted intelligent reflecting surface symbiotic radio system. In the system, the UAV acts as a relay for the intelligent reflecting surface to assist it in information transmission.

[0012] In summary, although the use of IRS has brought new development opportunities for solving the physical layer security of drone communication systems, there are still some problems that need to be solved. Since the smart reflective surface can change the amplitude or phase of the incident signal, fine-grained beamforming can be achieved.

[0013] However, as a hardware, the intelligent reflector cannot achieve continuous phase change due to the current hardware, and can only provide discrete phase shift. At present, some teams can achieve 1-bit phase deflection (that is, 180 degrees or no deflection), and the phase adjustment affects the accuracy of beamforming. How to improve the phase adjustment technology, make the phase adjustment closer to the actual UAV system and improve the beamforming accuracy, thereby improving the physical layer security performance of the system is worthy of further study. Summary of the invention

[0014] The purpose of the present invention is to address the deficiencies of the above-mentioned prior art and provide a UAV secure communication method based on IRS, which intends to combine intelligent reflection surface beamforming and phase adjustment to improve the security performance of the system physical layer by obtaining the optimal solution of beamforming and phase adjustment.

[0015] In order to solve the above technical problems, the present invention adopts the following technical solutions to achieve the above problems:

[0016] The present invention proposes a UAV secure communication method based on IRS, which coordinates through IRS deployed between two communicating UAVs A and B. The IRS is carried on UAV C, and there is an eavesdropping UAV E in the communication system. By adjusting the phase of the signal sent by UAV A to align it with UAV B, the signal forwarding process is completed, thereby improving the security rate of the system;

[0017] The safety rate R of the communication system is related to the beamforming vector w and the IRS phase θ. The method of alternately optimizing the beamforming vector w and the IRS phase θ is adopted to obtain the optimal beamforming vector w and the IRS phase θ in turn, thereby calculating the maximum value of the system safety rate R.

[0018] In some embodiments, the transmission process of the communication system includes the following steps:

[0019] S1: Drone A broadcasts a signal, and the signal received by the smart reflective surface is:

[0020]

[0021] Among them, x i is the information sequence sent at time i, P is the transmission power of drone A, n' AI The mean received by IRS is 0 and the variance is Additive Gaussian white noise;

[0022] S2: IRS receives and forwards, and the signals received by drone B and eavesdropping drone E are:

[0023]

[0024] in,

[0025] is the reflection phase diagonal matrix, β∈[0,1] is the reflection surface amplitude, θ∈[0,2π] is the reflection surface phase shift, n B 、n E The mean is 0 and the variances are Additive Gaussian white noise, x i =ws(t) is the message sequence sent by UAV A;

[0026] S3: Calculate the security rate R of the communication system. The security rates of drone B and eavesdropping drone E are:

[0027]

[0028]

[0029] Therefore, the safe rate R of the communication system is:

[0030]

[0031] Where w∈C m×1 is the beamforming vector, s(t) is the signal broadcast at time t, and ||w|| 2 ≤P max , E{|s(t)| 2}=1,P max is the maximum transmission power of UAV A.

[0032] In some implementations, in order to improve the safe transmission of the UAV channel, the maximum value of the safe rate R of the communication system is obtained as follows:

[0033]

[0034] It can be seen from the above formula that the safety rate is related to θ and w, and the objective function is non-convex with respect to variables θ and w. Therefore, the method of alternating optimization to achieve the optimal solution is used to maximize the objective function.

[0035] In some implementations, the method of alternately optimizing the beamforming vector w and the IRS phase θ includes the following process:

[0036] Step 1: Input the beamforming vector w, IRS phase θ and the transmission power P of UAV A into the simulation data model;

[0037] Step 2: Given the IRS phase θ, optimize the beamforming vector w;

[0038] Step 3: Given the beamforming vector w, optimize the IRS phase θ;

[0039] Step 4: Find the fastest descending direction of the gradient of the Riemann submanifold M;

[0040] Step 5: Calculate the system safety rate R using the optimal beamforming vector w and IRS phase θ to obtain the maximum value of the system safety rate R.

[0041] In some implementations, when optimizing the beamforming vector w in step 2, assuming that the IRS phase θ is constant,

[0042]

[0043]

[0044] Then formula (7) can be transformed into:

[0045]

[0046] According to the Rayleigh-Ritz theorem, the optimal solution of formula (22) is located at the characteristic vector (X B ,X E ) direction, and the power satisfies the normalization constraint ||w|| 2 =1, therefore, the optimal beamforming vector is:

[0047]

[0048] Where I is the N*N identity matrix.

[0049] In some implementations, when optimizing the IRS phase θ in step 3, assuming that the beamforming vector w is constant, let Then formula (7) can be transformed into:

[0050]

[0051] because is a constant, so formula (8) can be simplified to:

[0052]

[0053] The Euclidean gradient of the objective function is obtained through mathematical programming and Riemann flow optimization, and the optimal solution of the IRS phase θ is obtained according to the gradient descent method.

[0054] In some implementations, since the above formula (9) satisfies the conditions of fractional mathematical programming, it can be further transformed into:

[0055] Among them, st|θ i |=1,2,...,N,

[0056]

[0057] make

[0058] Where X=(|y1| 2 +|y2| 2 )γγ H , R[.] is the real part operation, |θ i |=1,2,...,N, y'1, y'2 represent the optimal solutions of y1, y2;

[0059] From formula (11), we can see that H is a semi-positive definite matrix, so f2(Φ) is a quadratic concave function with respect to Φ.

[0060] In some embodiments, Riemann flow optimization is used to solve f2(Φ), and the Euclidean gradient of the objective function is:

[0061] Δf2(Φ)=2XΦ (13),

[0062] Point θ k The tangent space at is:

[0063]

[0064] The Riemann gradient of the objective function is the projection on the Euclidean gradient space. The Riemann gradient gradf2(θ k ) is expressed as

[0065] I is the N×N identity matrix. According to the gradient descent method, the search direction in the tangent space is:

[0066] η=-gradf2(θ k ) (16),

[0067] And the space is updated as follows:

[0068] θ k+1 =θ k -ρgradf2(θ k ) (17),

[0069] ρ is the step size, and the point θ obtained at this time k+1 is not on the original manifold Ψ, so it needs to be mapped to the original manifold, that is,

[0070] θ k+1 =θ k+1 / ||θ k+1 || (18),

[0071] The optimal solution can be obtained by using the three formulas:

[0072] θ' k+1 =θ k +ρη (19).

[0073] In some embodiments, the IRS is a two-dimensional IRS composed of M reflective elements, and the IRS is placed between two communicating drones A and B and used as a passive relay;

[0074] And / or, the IRS is controlled by a small controller via an independent control link.

[0075] In some embodiments, the channels of drone A, drone B and eavesdropping drone E are all Rayleigh fading channels.

[0076] The beneficial effects achieved by the present invention are:

[0077] (1) Compared with the prior art, the present invention uses an intelligent reflective surface deployed on the relay drone to reflect as many signals as possible to the legitimate drones, while reducing the number of signals received by illegal drones, thereby improving the safety rate of the system.

[0078] (2) As a hardware, the intelligent reflective surface cannot realize continuous phase change, but can only provide discrete phase offset, and the phase adjustment will directly affect the accuracy of beamforming. The present invention adopts a solution combining the beamforming and phase adjustment of the intelligent reflective surface, obtains the optimal solution of beamforming and phase adjustment, improves the phase adjustment technology, makes the phase adjustment closer to the actual UAV system and improves the beamforming accuracy, thereby improving the physical layer security performance of the system.

[0079] (3) The system safety rate obtained by the joint optimization IRS relay method in the present invention is higher than that of the case where traditional AF communication is used and no relay is used; at the same time, the use of IRS relay can effectively reduce system power consumption and save the system's sending function. BRIEF DESCRIPTION OF THE DRAWINGS

[0080] Figure 1 A model diagram of a communication system provided by an embodiment of the present invention;

[0081] Figure 2 A safety speed comparison diagram of an embodiment of the present invention and a comparative example;

[0082] Figure 3 This is a comparison chart of power consumption under the same confidentiality rate conditions between the embodiment of the present invention and the comparative example. DETAILED DESCRIPTION

[0083] In order to facilitate the understanding of those skilled in the art, the present invention is further described below in conjunction with embodiments and drawings. The contents mentioned in the implementation modes are not intended to limit the present invention.

[0084] Since the intelligent reflective surface is passive, low-power, and intelligently controllable, its development has brought new ideas to the physical layer security issues of UAV systems. The intelligent reflective surface can change the amplitude or phase of the incident signal and can achieve fine-grained beamforming; however, as a hardware, the phase adjustment of the intelligent reflective surface is not continuous. Therefore, the present invention studies the joint optimization of beamforming and phase adjustment schemes for UAV communication systems with eavesdropping nodes. By deploying the intelligent reflective surface on the relay UAV, the signal is reflected to the legal UAV as much as possible while reducing the signal received by the illegal UAV, thereby improving the security rate of the system. By comparing the security rates obtained by the system using different relay modes and the system power consumed under the condition of achieving the same security rate, the simulation results show that the system using the intelligent reflective surface as the relay node can effectively improve the system security rate and reduce the system power consumption.

[0085] Depend on Figure 1 It can be seen that the eavesdropping drone E has a high correlation with the user drone B. If a common relay node is used, the eavesdropping drone E has a high probability of eavesdropping information, and can even achieve the same safety rate as the user drone B. Therefore, in this embodiment, an IRS-based drone security communication method coordinates through an IRS deployed between two communicating drones A and drone B. The IRS is mounted on drone C. There is an eavesdropping drone E in the communication system. By adjusting the phase of the signal sent by drone A to align it with drone B, the signal forwarding process is completed, and the safety rate of the system is improved.

[0086] The safety rate R of the communication system is related to the beamforming vector w and the IRS phase θ. The method of alternately optimizing the beamforming vector w and the IRS phase θ is adopted to obtain the optimal beamforming vector w and the IRS phase θ in turn, thereby calculating the maximum value of the system safety rate R.

[0087] The transmission process of the above communication system includes the following steps:

[0088] S1: Drone A broadcasts a signal, and the signal received by the smart reflective surface is:

[0089]

[0090] Among them, x i is the information sequence sent at time i, h AI is the channel coefficient for sending UAV A to IRS, P is the transmission power of UAV A, n' AI The mean received by IRS is 0 and the variance is Additive Gaussian white noise;

[0091] S2: IRS receives and forwards the signal. The signals received by drone B and eavesdropping drone E are:

[0092]

[0093] in, is the reflected phase diagonal matrix, is the channel coefficient from IRS to the receiving drone B and the eavesdropping drone;

[0094] β∈[0,1] is the amplitude of the reflection surface, θ∈[0,2π] is the phase shift of the reflection surface, n B 、n E The mean is 0 and the variances are Additive Gaussian white noise, x i =ws(t) is the message sequence sent by UAV A.

[0095] S3: Calculate the security rate R of the communication system. The security rates of drone B and eavesdropping drone E are:

[0096]

[0097]

[0098] Therefore, the safe rate R of the communication system is:

[0099]

[0100] Where w∈C m×1 is the beamforming vector, s(t) is the signal broadcast at time t, and ||w|| 2 ≤P max , E{|s(t)| 2}=1,P max is the maximum transmission power of UAV A.

[0101] In order to improve the safe transmission of UAV channels, the maximum value of the safe rate R of the communication system is obtained as:

[0102]

[0103] It can be seen from the above formula that the safety rate is related to θ and w, and the objective function is non-convex with respect to variables θ and w. Therefore, the method of alternating optimization to achieve the optimal solution is used to maximize the objective function.

[0104] The method of alternately optimizing the beamforming vector w and the IRS phase θ includes the following steps:

[0105] Step 1: Input the beamforming vector w, IRS phase θ and the transmission power P of UAV A into the simulation data model;

[0106] Step 2: Given the IRS phase θ, optimize the beamforming vector w;

[0107] Step 3: Given the beamforming vector w, optimize the IRS phase θ;

[0108] Step 4: Find the fastest descending direction of the gradient of the Riemann submanifold M;

[0109] Step 5: Calculate the system safety rate R using the optimal beamforming vector w and IRS phase θ to obtain the maximum value of the system safety rate R.

[0110] When optimizing the beamforming vector w in step 2, assuming that the IRS phase θ is constant,

[0111]

[0112]

[0113] Then formula (7) can be transformed into:

[0114]

[0115] st||w|| 2 =1 (22),

[0116] Depend on The Rayleigh-Ritz theorem shows that the optimal solution of formula (22) is located at the characteristic vector (X B ,X E ) direction, and the power satisfies the normalization constraint ||w|| 2 =1, therefore, the optimal beamforming vector is:

[0117]

[0118] Where I is the N*N identity matrix.

[0119] When optimizing the IRS phase θ in step 3, assuming that the beamforming vector w is constant, let Then formula (7) can be transformed into:

[0120]

[0121] because is a constant, so formula (8) can be simplified to:

[0122]

[0123] The Euclidean gradient of the objective function is obtained through mathematical programming and Riemann flow optimization, and the optimal solution of the IRS phase θ is obtained according to the gradient descent method.

[0124] Since the above formula (9) meets the conditions of fractional mathematical programming, it can be further transformed into:

[0125]

[0126] in,

[0127]

[0128] make

[0129] Where X=(|y1| 2 +|y2| 2 )γγ H , R[.] is the real part operation, |θ i |=1,2,...,N, y'1, y'2 represent the optimal solutions of y1, y2;

[0130] From formula (11), we can see that H is a semi-positive definite matrix, so f2(Φ) is a quadratic concave function with respect to Φ.

[0131] Riemann flow optimization is used to solve f2(Φ), and the Euclidean gradient of the objective function is:

[0132] Δf2(Φ)=2XΦ (13),

[0133] Point θ k The tangent space at is:

[0134]

[0135] The Riemann gradient of the objective function is the projection on the Euclidean gradient space. The Riemann gradient gradf2(θ k ) is expressed as

[0136] I is the N×N identity matrix. According to the gradient descent method, the search direction in the tangent space is:

[0137] η=-gradf2(θ k ) (16),

[0138] And the space is updated as follows:

[0139] θ k+1 =θ k -ρgradf2(θ k ) (17),

[0140] ρ is the step size, and the point θ obtained at this time k+1is not on the original manifold Ψ, so it needs to be mapped to the original manifold, that is,

[0141] θ k+1 =θ k+1 / ||θ k+1 || (18),

[0142] The optimal solution can be obtained by using the three formulas:

[0143] θ' k+1 =θ k +ρη (19).

[0144] In this embodiment, the IRS is a two-dimensional IRS composed of M reflective elements. The IRS is placed between two communicating drones A and B and used as a passive relay. The IRS is controlled by a small controller through an independent control link.

[0145] Since the model in this embodiment is a UAV communication system, the channel can be modeled as a Rayleigh channel, and the channels of UAV A, UAV B, and eavesdropping UAV E are all Rayleigh fading channels. In order to reflect the advantages of the IRS relay proposed in this embodiment in improving the system security rate, the security rates of no relay node (Comparative Example 1), AF (Amplify-and-Forward) relay node (Comparative Example 2) and IRS forwarding node (this embodiment) will be compared respectively, and the transmission power of the transmitter required by different relays under the condition of achieving the same security rate will be analyzed.

[0146] Assuming that there are malicious eavesdropping drone nodes, the confidentiality rate of the legitimate channel based on IRS communication is analyzed, that is, the trend of formula (6) with power change. Since the communication nodes are all drone nodes, a two-dimensional Cartesian coordinate system is introduced, the position of drone A is [0; 100], the position of drone B is [150; 100], the position of eavesdropping drone E is [200; 150], and the position of drone C equipped with IRS is [50; 100]. From the simulation results (such as Figure 2 ) It can be seen that the safety rate of the communication system using the joint optimization of the IRS beamforming vector and phase is higher than that of the other two communication systems, and increases with the increase of the transmission power. Therefore, it can be proved that the UAV communication system based on IRS assistance can improve the safety performance of the system.

[0147] Since the reflection unit of IRS is equivalent to a mirror and does not require additional transmission power, the energy efficiency of the system can be improved. In order to better demonstrate the role of IRS in energy efficiency, this embodiment compares the transmission power required by relay nodes of different relay modes at the same height and different distances (50 meters to 15 meters) under the condition of achieving the same confidentiality rate (R = 5bps / Hz). From the simulation results (such as Figure 3) It can be seen that the power consumption of the system without relay nodes (Comparative Example 1) and with AF relay nodes (Comparative Example 2) increases with the increase of distance, while the power consumption of the IRS-assisted relay system (this embodiment) is relatively low and can reach the lowest within the range of 70-90 meters. Therefore, using IRS for relaying can effectively save the system's transmission power.

[0148] In this embodiment, the UAV communication system with IRS nodes as relays is analyzed. Due to the existence of eavesdropping nodes, the method of jointly optimizing beamforming vectors and phases is used to improve the physical layer security performance of the system. The simulation results show that the system security rate obtained by using the jointly optimized IRS relay method is higher than that of using traditional AF communication and not using relays. In addition, the transmission functions required by different relay methods are analyzed under the premise of achieving the same security rate. The simulation results show that using IRS relays can effectively reduce system power consumption and save the system's transmission function.

[0149] The above embodiments are preferred implementation schemes of the present invention. In addition, the present invention may also be implemented in other ways. Any obvious replacement without departing from the concept of the present technical solution is within the protection scope of the present invention.

[0150] In order to make it easier for ordinary technicians in the field to understand the improvements of the present invention over the prior art, some drawings and descriptions of the present invention have been simplified, and for the sake of clarity, some other elements are omitted in this application document. Ordinary technicians in the field should realize that these omitted elements may also constitute the content of the present invention.

Claims

1. A secure communication method for unmanned aerial vehicles based on IRS, characterized in that: Coordination is performed through the IRS deployed between two communicating drones A and B. The IRS is mounted on drone C. There is an eavesdropping drone E in the communication system. By adjusting the phase of the signal sent by drone A to align it with drone B, the signal forwarding process is completed, thereby improving the security rate of the system; The safety rate R of the communication system is related to the beamforming vector w and the IRS phase θ. The method of alternately optimizing the beamforming vector w and the IRS phase θ is adopted to obtain the optimal beamforming vector w and the IRS phase θ in turn, thereby calculating the maximum value of the system safety rate R.

2. The IRS-based UAV secure communication method according to claim 1, characterized in that: The transmission process of the communication system comprises the following steps: S1: Drone A broadcasts a signal, and the signal received by the smart reflective surface is: Among them, x i is the information sequence sent at time i, h AI is the channel coefficient from UAV A to IRS, P is the transmission power of UAV A, The mean received by IRS is 0 and the variance is Additive Gaussian white noise; S2: IRS receives and forwards the signals received by drone B and eavesdropping drone E respectively: in, is the reflected phase diagonal matrix, is the channel coefficient from IRS to the receiving drone B and the eavesdropping drone, β∈[0,1] is the amplitude of the reflection surface, θ∈[0,2π] is the phase shift of the reflection surface, n B 、n E The mean is 0 and the variances are Additive Gaussian white noise, x i =ws(t) is the message sequence sent by drone A; S3: Calculate the security rate R of the communication system. The security rates of drone B and eavesdropping drone E are: Therefore, the safe rate R of the communication system is: Where w∈C m×1 is the beamforming vector, s(t) is the signal broadcast at time t, and ||w|| 2 ≤P max , E{|s(t)| 2 }=1,P max is the maximum transmission power of UAV A.

3. The IRS-based drone secure communication method according to claim 2, characterized in that: In order to improve the safe transmission of UAV channels, the maximum value of the safe rate R of the communication system is obtained as: S.1.|w| 2 ≤P m3X |θ i |=1,i=1,2,...,N It can be seen from the above formula that the safety rate is related to θ and w, and the objective function is non-convex with respect to variables θ and w. Therefore, the method of alternating optimization to achieve the optimal solution is used to maximize the objective function.

4. The IRS-based drone secure communication method according to claim 3, characterized in that: The method of alternately optimizing the beamforming vector w and the IRS phase θ includes the following steps: Step 1: Input the beamforming vector w, IRS phase θ and the transmission power P of UAV A into the simulation data model; Step 2: Given the IRS phase θ, optimize the beamforming vector w; Step 3: Given the beamforming vector w, optimize the IRS phase θ; Step 4: Find the fastest descending direction of the gradient of the Riemann submanifold M; Step 5: Calculate the system safety rate R using the optimal beamforming vector w and IRS phase θ to obtain the maximum value of the system safety rate R.

5. The IRS-based UAV secure communication method according to claim 4, characterized in that: When optimizing the beamforming vector w in step 2, assuming that the IRS phase θ is constant, Then formula (7) can be transformed into: in According to the Rayleigh-Ritz theorem, the optimal solution of formula (22) is located at the characteristic vector (X B ,X E ) direction, and the power satisfies the normalization constraint ||w|| 2 =1, therefore, the optimal beamforming vector is: Where I is the N*N identity matrix.

6. The IRS-based UAV secure communication method according to claim 5, characterized in that: When optimizing the IRS phase θ in step 3, assuming that the beamforming vector w is constant, let Then formula (7) can be transformed into: because is a constant, so formula (8) can be simplified to: The Euclidean gradient of the objective function is obtained through mathematical programming and Riemann flow optimization, and the optimal solution of the IRS phase θ is obtained according to the gradient descent method.

7. The IRS-based UAV secure communication method according to claim 6, characterized in that: Since the above formula (9) meets the conditions of fractional mathematical programming, it can be further transformed into: where s.t. |θ i | = 1, 2, ..., N, make Where X=(|y1| 2 +|y2| 2 )γγ H , R[.] is the real part operation, |θ i |=1,2,...,N, y'1, y'2 represent the optimal solutions of y1, y2; From formula (11), we can see that H is a semi-positive definite matrix, so f2(Φ) is a quadratic concave function with respect to Φ.

8. The IRS-based UAV secure communication method according to claim 7, characterized in that: Riemann flow optimization is used to solve f2(Φ), and the Euclidean gradient of the objective function is: Δf2(Φ)=2XΦ (13), Point θ k The tangent space at is: The Riemann gradient of the objective function is the projection on the Euclidean gradient space. The Riemann gradient gradf2(θ k ) is expressed as I is the N×N identity matrix. According to the gradient descent method, the search direction in the tangent space is: η=-gradf2(θ k ) (16), And the space is updated as follows: i k+1 =θ k -ρgradf2(θ k ) (17), ρ is the step size, and the point θ obtained at this time k+1 is not on the original manifold Ψ, so it needs to be mapped to the original manifold, that is, i k+1 =θ k+1 / ||θ k+1 || (18), The optimal solution can be obtained by using the three formulas: I will k+1 =θ k +ρη (19).

9. The IRS-based drone secure communication method according to any one of claims 1 to 8, characterized in that: The IRS is a two-dimensional IRS composed of M reflective elements, which is placed between two communicating drones A and B and used as a passive relay; And / or, the IRS is controlled by a small controller via an independent control link.

10. The IRS-based UAV secure communication method according to claim 1, characterized in that: The channels of the drone A, drone B and eavesdropping drone E are all Rayleigh fading channels.

Images