Intelligent reflecting surface assisted unmanned aerial vehicle safety communication and perception integrated design method

Abstract

The application provides a kind of unmanned aerial vehicle safety communication perception integrated design method assisted by intelligent reflecting surface, belongs to the communication perception integrated field of unmanned aerial vehicle communication network.Unmanned aerial vehicle is used as air base station, transmits communication signal and perception signal to realize the function of communication perception integration, while ensuring safety.Based on the idea of alternating optimization, the original problem is decomposed into four sub-problems for solving, and the non-convex optimization problem is converted into a solvable convex optimization problem using continuous convex approximation technology.The application provides a design method for joint power allocation, reflecting surface phase shift, user and target timing arrangement, and unmanned aerial vehicle flight trajectory, which can generate the optimal network design according to the model parameters to maximize the average achievable rate.

Description

Technical Field

[0001] This invention belongs to the field of communication and sensing integration in UAV communication networks. It relates to a design scheme that combines a UAV with an intelligent reflective surface to achieve secure communication and sensing integration. Specifically, when achieving communication and sensing integration, the following three aspects are jointly optimized: transmission power allocation, UAV trajectory and speed, and timing arrangement of users and targets, thereby maximizing the achievable rate of the system. Background Technology

[0002] Communication-sensing integration is a promising technology for reducing spectrum congestion, allowing sensing and communication to share the same spectrum and hardware. Compared to separate radar and communication systems, communication-sensing integration offers advantages such as smaller size, lower power consumption, and less mutual interference. Therefore, it has a wide range of applications, including connected vehicles and the Internet of Things (IoT). Furthermore, communication-sensing integration has attracted considerable research interest from many researchers.

[0003] On the other hand, intelligent reflective surfaces (IRS) can reconfigure the wireless propagation environment by controlling the amplitude and phase of the reflected signal. Specifically, an intelligent reflective surface (IRS) is a metasurface composed of a large number of passive reflective elements. Furthermore, intelligent reflective surfaces are passive. Therefore, compared to active devices, intelligent reflective surfaces have the advantages of simple structure, low cost, low power consumption, and flexible deployment. Intelligent reflective surfaces can not only reconfigure wireless channels but also provide virtual line-of-sight links for target sensing.

[0004] Unmanned aerial vehicles (UAVs) can serve as relays in communication systems to extend transmission range. However, UAV relay communication increases information transmission latency, and the energy consumed by the UAV during information processing and its own movement also consumes significant propulsion energy. Considering the limited energy carrying capacity and endurance of UAVs, improving system energy efficiency and achieving green communication is a key issue. The literature [S.Li,B.Duo,X.Yuan,Y.-C.Liangand M.Di Renzo, "Reconfigurable Intelligent Surface Assisted UAV Communication: Joint Trajectory Design and Passive Beamforming," in IEEE Wireless Communications Letters, vol.9, no.5, pp.716-720, May 2020, doi:10.1109 / LWC.2020.2966705.] proposes an intelligent reflective surface-assisted UAV communication scheme, but it only considers a single user and is difficult to extend to multi-user scenarios. With the increasing number of terminals, UAVs often serve more than one user; therefore, research is needed on UAV communication systems for user scenarios.

[0005] This invention combines the many advantages of intelligent reflectors, integrating UAVs with passive intelligent reflectors to assist communication. This not only saves energy but also significantly improves the quality and speed of space communication. Aiming to maximize the average reachability, this invention appropriately designs the system parameters, thereby generating an optimal network design based on model parameters to maximize the average reachability. Summary of the Invention

[0006] The purpose of this invention is to solve the problem of maximizing the average reachable rate of an air-to-ground wireless network for unmanned aerial vehicles (UAVs) assisted by a smart reflector. In the network model, the UAV acts as a dual-function base station, serving K users and sensing J targets. A potential eavesdropper exists in the network, attempting to intercept users' confidential information. A specific solution is illustrated in the diagram. Figure 1 As shown in the figure. Based on this model, the present invention provides a design method for jointly optimizing transmit power allocation, user and target timing, passive beamforming of intelligent reflectors, and UAV trajectory, in order to maximize the average achievable rate of the system.

[0007] To achieve the above-mentioned technical objectives, the technical solution adopted by the present invention is as follows:

[0008] A design method for an integrated security communication and sensing system for unmanned aerial vehicles (UAVs) assisted by an intelligent reflective surface includes the following steps:

[0009] The first step is to build a system model:

[0010] (1) The UAV is used as a dual-function base station to communicate with K ground users and sense J ground targets. Simultaneously, a potential eavesdropper is present to eavesdrop on the users' information, and the eavesdropper's channel state information cannot be obtained. The base station has a single antenna, and each user has a single antenna. In a three-dimensional coordinate system, the horizontal coordinates of the k-th (k=1,...,K) user and the j-th (j=1,...,J) target are u... k =(x u,k ,y u,k ) T and t j =(x t,j ,y t,j ) T The horizontal coordinate of the eavesdropper is p. e =(x e ,y e ) T ;

[0011] (2) Assume the drone travels at a fixed altitude H u Flight, with a flight time of T, is discretized into N time slots, with a time slot length of... Then the horizontal coordinate of the UAV in the nth time slot is q[n] = (q x [n],q y [n]) T From the initial position q I Fly to the destination position q F That is, q[0] = q I ,q[N]=q F ;

[0012] (3) Assuming the channel between the base station and the user is a line-of-sight channel, then the channel between the UAV and user k, target j and the eavesdropper in the nth time slot is:

[0013]

[0014] Where L0 represents the channel gain per unit reference distance, H represents the distance between the UAV and ground user k in the nth time slot. u This represents the drone's altitude; α is the path loss index. Let t be the distance between the UAV and target j in the nth time slot. j This represents the horizontal coordinate of the j-th (j = 1, ..., J) target; Let p be the distance between the drone and the eavesdropper in the nth time slot. e The horizontal coordinates of the eavesdropper;

[0015] The channel gain from the UAV to the smart reflector is expressed as:

[0016]

[0017] Where κ represents the path loss exponent. a(θ) represents the distance between the base station and the smart reflector. ai [n],φ ai [n]) represents the steering vector at time slot n, where θ ai [n] represents the pitch angle in the nth time slot, φ ai [n] represents the azimuth angle of the nth time slot, and the specific expression is shown in formula (3):

[0018]

[0019] in, This represents the phase delay of the steering vector along the x-axis, x i q represents the x-axis coordinate value of the intelligent reflective surface. x [n] represents the x-axis coordinate value of the UAV in the nth time slot, d ai [n] represents the distance between the UAV and the reflector in the nth time slot; The y-axis represents the phase delay of the steering vector; λ represents the wavelength; d represents the element spacing; M x M represents the number of units on the x-axis of the reflecting surface. y This indicates the number of units on the y-axis of the reflecting surface; Indicates the imaginary part unit. It represents the Kronecker product.

[0020] The channel from the smart reflector to the ground user is also a line-of-sight link, and its gain is expressed as:

[0021]

[0022] in, This indicates the distance between the smart reflective surface and the user. This indicates the phase delay of the steering vector along the x-axis. This indicates the phase delay of the steering vector along the y-axis. Denotes the Kronecker product, x u,k This represents the x-axis coordinate of the k-th user;

[0023] The phase shift adjustment of the signal on the smart reflector can be expressed as: Where diag represents a diagonal matrix, θ m [n](θ∈[0,2π]) is the phase shift value of the m-th reflection unit in the n-th time slot;

[0024] (4) Assuming that the UAV uses time division multiple access to eliminate multi-user interference, that is, the UAV communicates with only one user in a specific time slot, we have the following constraints:

[0025]

[0026] If the drone serves user k in the nth time slot, α k [n] = 1; if the drone serves other users in the nth time slot, α k [n] = 0, where α k [n] represents the associated indicator variable for the k-th user in the n-th time slot; k = 1, ..., K represents the user index;

[0027] (5) Assume that to reduce the computational complexity of parameter estimation, the UAV will detect at most one target within a specific time slot. To ensure perception performance, the j-th target should be detected at least N times. j Therefore, we have the following constraints:

[0028]

[0029] If the UAV detects target j in the nth time slot, β j [n] = 1; if the UAV detects other targets in the nth time slot, β j [n] = 0, β j [n] represents the associated indicator variable for the j-th target in the n-th time slot; j = 1, ..., J represents the target index;

[0030] (6) The transmitted signal in the nth time slot is:

[0031]

[0032] in, This represents the communication symbol of the k-th user; Indicates a sensed signal; p c [n] and p r [n] represents the power allocated to the communication and sensing signals, respectively; K represents the number of users; J represents the number of targets; This is a complex Gaussian distribution with a mean of 0 and a variance of 1.

[0033] (7) Assume that each user uses serial interference cancellation to eliminate interference from the sensed signal, meaning each user removes the sensed signal before decoding its own signal. To ensure serial interference at each user, we have the following constraints:

[0034]

[0035] in, and Let represent the equivalent channels from the UAV in the nth time slot to user k and target j, respectively; This represents the channel from the UAV in the nth time slot to user k; Φ[n] represents the channel from the reflector to user k; Φ[n] represents the phase shift matrix of the nth time slot reflector; h ai [n] represents the channel from the UAV in the nth time slot to the reflector; This represents the channel from the UAV in the nth time slot to target j; This represents the channel from the UAV to the target in the nth time slot;

[0036] (8) The reachable rate of user k in the nth time slot can be expressed as:

[0037]

[0038] Where, σ 2 Indicates noise power; α k [n] represents the associated indicator variable for the k-th user in the n-th time slot; p c [n] represents the power allocated to the communication signal in the nth time slot; h k [n] represents the equivalent channel from the UAV in the nth time slot to user k.

[0039] (9) The received power of the j-th target in the n-th time slot can be expressed as:

[0040]

[0041] p c [n] and p r [n] represents the power allocated to the communication and sensing signals, respectively; This represents the equivalent channel from the UAV in the nth time slot to the target j;

[0042] (10) Assuming the eavesdropper does not employ serial interference cancellation, the eavesdropping rate in the nth time slot can be expressed as:

[0043]

[0044] in, h represents the equivalent channel between the drone and the eavesdropper in the nth time slot; ae [n] represents the channel from the drone to the eavesdropper in the nth time slot; h ie This represents the channel from the nth time slot reflector to the eavesdropper.

[0045] (11) The average safety rate can be expressed as:

[0046]

[0047] in,[·] + =max(·,0),R k [n] represents the communication rate of user k in the nth time slot; R e [n] represents the eavesdropping rate of the nth time slot.

[0048] The second step is to determine the objective function and optimization variables, and to list the optimization problem:

[0049] By analyzing the trajectory {q[n]}, speed {v[n]}, and power allocation {p} of the base station drone... c [n]} and {p r [n]}, the phase shift matrix of the intelligent reflector {Φ[n]}, user scheduling {α k [n]} and target scheduling {β j Joint optimization of [n]} leads to the following optimization problem.

[0050]

[0051] In this optimization problem, P max Indicates the maximum instantaneous transmit power. Γ represents the average transmit power. j [n] represents the power threshold of the j-th target in the n-th time slot, C1-C2 represent power constraints, C3-C4 represent serial interference cancellation constraints, C5 represents binary constraints, C9 represents received power constraints, C10-C11 represent trajectory constraints, and C13-C14 represent velocity constraints.

[0052] The third step is to design an algorithm to solve the optimization problem:

[0053] Using the idea of ​​block iteration, the optimization problem shown in the above formula (13) is decomposed into four sub-problems, and the power allocation problem is solved using CVX; for UAV trajectory optimization, the non-convex problem is approximately transformed into a convex optimization problem by applying the continuous convex approximation method; for the reflector phase optimization problem, the Riemannian manifold optimization and penalty term method is used for solution; for the target and user timing problem, the penalty function method is used for solution; the specific steps are as follows:

[0054] (1) Optimize power allocation {p c [n],p r [n]}

[0055] Given {Φ[n], α k [n],β j [n],p c [n],p r The optimization problem is represented as [n]}.

[0056]

[0057] This problem is a convex optimization problem, which can be solved using CVX.

[0058] (2) Optimize the trajectory {q[n]} and velocity {v[n]}

[0059] Given {Φ[n], α k [n],β j [n],p c [n],p r The optimization problem is represented as [n]}.

[0060]

[0061] This problem is difficult to solve because of the existence of a non-concave objective function and non-convex constraints. The trajectory {q[n]} exists not only in the distance term but also in h. ai In the guiding vector of [n]. And, h ai The guiding vector of [n] is a complex function of the trajectory {q[n]}, which makes the problem difficult to solve. To address this issue, a continuous convex approximation with a trust region is used to approximate h. ai The guiding vector in [n], i.e.

[0062]

[0063] in, This represents the trajectory obtained in the previous iteration. Let be the distance from the UAV to the reflector in the nth time slot;

[0064] To ensure approximate accuracy, we have the following constraints.

[0065]

[0066] Where δ represents the precision threshold;

[0067] R k [n] can be represented as

[0068]

[0069] in, This represents the distance vector between the UAV in the nth time slot and user k.

[0070]

[0071] and This represents the approximate channel between the UAV and the reflector in the nth time slot.

[0072] Similarly, It can be represented as

[0073]

[0074] in, This represents the approximate channel between the UAV in the nth time slot and the target j. This represents the distance vector between the UAV in the nth time slot and the target j.

[0075]

[0076] It is an Hermitian matrix, where p[n] = p c [n]+p r [n] represents the transmit power of the UAV in the nth time slot.

[0077] Therefore, the optimization problem can be expressed as

[0078]

[0079] By introducing slack variable {γ k [n]}, which satisfies

[0080]

[0081] The optimization problem can be equivalently represented as

[0082]

[0083] By introducing slack variables and Its satisfaction

[0084]

[0085] The optimization problem (22) can be equivalently represented as

[0086]

[0087] Furthermore, by substituting q[n] into... d ai [n] and (25) can be equivalently expressed as

[0088]

[0089] These constraints are in the form of convexity. By continuous convex approximation, (26) can be transformed into a convex problem as shown in (28).

[0090] stC1, C2, C3, C9

[0091]

[0092]

[0093]

[0094]

[0095]

[0096]

[0097]

[0098] (28) It can be solved using CVX.

[0099] (3) Optimize the phase of the reflecting surface {Φ[n]},

[0100] Given {p c [n],p r [n],q[n],v[n],α k [n],β j The optimization problem for [n] can be represented as:

[0101]

[0102] This is a non-convex optimization problem, which is very difficult to solve.

[0103] definition p c [n]|h k [n]| 2 and It can be equivalently represented as

[0104]

[0105] in, Represents the augmented phase shift vector; Let represent the equivalent channel matrix from the UAV in the nth time slot to user k; Let represent the equivalent channel matrix from the UAV in the nth time slot to the target j;

[0106]

[0107] Therefore, the optimization problem can be equivalently represented as

[0108]

[0109] Furthermore, we can handle the above problems in parallel. Specifically, the phase optimization problem in the nth time slot can be expressed as:

[0110]

[0111] By moving constraints C1 and C2 from problem (30) to the objective function, the optimization problem can be expressed as follows:

[0112]

[0113] in,

[0114]

[0115] in, This represents the penalty coefficient. Because of the constraints in (34) It is a complex circular manifold, which can be solved using the Riemann conjugate gradient algorithm.

[0116] (4) Optimize {α k [n]} and {β j [n]}

[0117] Given {p c [n],p r The optimization problem is represented as [n],q[n],v[n],Φ[n]}.

[0118]

[0119] This problem is difficult to solve due to the presence of binary variables. Fortunately, C5 can be equivalently represented as

[0120]

[0121] Despite α k [n] and β j [n] is transformed into a continuous variable, constrained. and They are still non-convex. To solve this problem, we use a continuous convex approximation to transform them into the following convex form:

[0122]

[0123] By adding constraint (38) as a penalty term to the objective function of (36), problem (36) can be rewritten as follows:

[0124]

[0125] Here, γ > 0 represents the penalty coefficient. This problem is a convex optimization problem, which can be solved using CVX.

[0126] (5) The above optimization problems (referring to formulas 14, 28, 34, and 39) are solved using an alternating optimization algorithm. In each iteration, the power allocation problem is solved using CVX, followed by the UAV trajectory optimization problem using a continuous convex approximation algorithm. Then, the user and target scheduling problem is solved using continuous convex approximation and penalty functions. Finally, the Riemannian manifold optimization problem is used to solve the reflector phase optimization problem. The parameter values ​​are updated for the next iteration until the algorithm converges. The specific process is as follows:

[0127] 1) Set the initial power allocation {p c,0 [n],p r,0 [n]}, intelligent reflector reflection phase {Φ0[n]}, user scheduling {α k,0 [n]}, target scheduling {β j,0 [n]}, UAV flight trajectory {q0[n]}, UAV flight speed {v0[n]}, iteration count t=0;

[0128] 2) Solve the convex optimization problem (14) to obtain the power allocation result of the t-th iteration, i.e., {p c,t [n],p r,t [n]};

[0129] 3) Solve the convex optimization problem (26) to obtain the trajectory and velocity optimization results for the t-th iteration, i.e., {q t [n],v t [n]};

[0130] 4) Solve the optimization problem (34) to obtain the optimization result of the reflector phase in the t-th iteration, i.e., {Φ t [n]};

[0131] 5) Solve the convex optimization problem (39) to obtain the optimization results of the user and target scheduling in the t-th iteration, i.e., {α k,t [n],β j,t [n]};

[0132] 6) Update t = t + 1; skip to step 2) for the next iteration optimization;

[0133] 7) Until convergence. The convergence condition shown is: reaching the maximum number of convergences or if the increase in the optimization objective value is less than the threshold ε0.

[0134] This paper considers an integrated system for safe communication and perception of unmanned aerial vehicles (UAVs) assisted by intelligent reflective surfaces. In order to maximize the average reachability of the system, power allocation, UAV trajectory and speed, phase shift of the reflective surface, and timing of users and targets are jointly optimized. The designed UAV trajectory can maximize the average reachability of the system while ensuring safety.

[0135] The beneficial effects of this invention are:

[0136] This invention provides a scheme to maximize the average reachability of the system while satisfying perception requirements by jointly optimizing power allocation, UAV trajectory and speed, phase shift of the reflector, and user and target timing, while ensuring communication security. This invention also provides a reference value selection method for achieving integrated secure communication and perception for UAVs. Attached Figure Description

[0137] Figure 1 This is a schematic diagram of an integrated network for safe communication and perception of unmanned aerial vehicles (UAVs) assisted by intelligent reflective surfaces.

[0138] Figure 2 This is a comparison of the drone trajectories of different optimization schemes when the flight time is 50 seconds.

[0139] Figure 3 shows the curves of the average achievable rate as a function of the maximum transmit power for different schemes.

[0140] Figure 4 shows the average achievable rate of different schemes as a function of the number of smart reflective surface units.

[0141] Figure 5 shows the curves of average reachability as a function of flight time for different schemes.

[0142] Figure 6 These are the user timing results of the proposed solution. Detailed Implementation

[0143] To better understand the above technical solution, a detailed analysis is provided below in conjunction with the accompanying drawings and specific implementation methods.

[0144] A design method for an integrated security communication and sensing system for unmanned aerial vehicles (UAVs) assisted by an intelligent reflective surface includes the following steps:

[0145] The first step is to build a system model:

[0146] Establish a secure communication and sensing integrated system model. The UAV acts as a dual-function base station to communicate with K ground users and sense J ground targets. At the same time, there is a potential eavesdropper who is eavesdropping on the users' information, and the channel state information of the eavesdropper cannot be obtained.

[0147] The second step is to determine the objective function and optimization variables, and to list the optimization problem:

[0148] Through joint optimization of power allocation {p c [n],p r [n]}, intelligent reflector reflection phase {Φ[n]}, user scheduling {α k [n]}、Target scheduling {β jGiven [n]}, the UAV flight trajectory {q[n]} and the UAV flight speed {v[n]}, maximize the average reachability of the system, forming the following optimization problem (13);

[0149] The third step is to design an algorithm to solve the optimization problem:

[0150] (1) Power allocation {p c [n],p r Optimization of [n]}

[0151] Fixed intelligent reflector reflection phase {Φ[n]}, user scheduling {α k [n]}、Target scheduling {β j Given [n]}, the UAV flight trajectory {q[n]}, and the UAV flight speed {v[n]}, solve for the power allocation {p}. c [n],p r [n]} Subproblem (14);

[0152] (2) Optimization of the phase {Φ[n]} of the intelligent reflector

[0153] Fixed power allocation {p c [n],p r [n]}、User scheduling {α k [n]}、Target scheduling {β j Given the UAV flight trajectory {q[n]} and the UAV flight speed {v[n]}, solve the subproblems of trajectory {q[n]} and speed {v[n]} (26);

[0154] (3) Optimization of UAV flight trajectory {q[n]} and UAV flight speed {v[n]}

[0155] Fixed power allocation {p c [n],p r [n]}, intelligent reflector reflection phase {Φ[n]}, user scheduling variable {α k [n]} and the target scheduling variable {β j [n]}, solve the subproblems of trajectory {q[n]} and velocity {v[n]} (34);

[0156] (4) User scheduling {α k [n]} and target scheduling {β j Optimization of [n]}

[0157] Fixed power allocation {p c [n],p r Given the intelligent reflector reflection phase {Φ[n]}, trajectory {q[n]}, and velocity {v[n]}, solve for the user scheduling {α}. k[n]} and the target scheduling variable {β j [n]} Subproblem (39);

[0158] Step 4: Design the iterative algorithm:

[0159] The optimization problem is solved using an alternating iterative algorithm. In each iteration, continuous convex approximation is used to solve the power allocation subproblem, trajectory and velocity optimization subproblem, and user and target scheduling subproblem. Riemannian manifold optimization is used to solve the reflector phase optimization subproblem, thereby maximizing the average reachability rate. Finally, the values ​​of the parameters are updated for the next iteration, until the algorithm converges. The specific process is as follows:

[0160] 1) Initialize power allocation {p c,0 [n],p r,0 [n]}, intelligent reflector reflection phase {Φ0[n]}, user scheduling {α k,0 [n]}, target scheduling {β j,0 [n]}, UAV flight trajectory {q0[n]}, UAV flight speed {v0[n]}, iteration count t=0, iteration termination threshold ε0;

[0161] 2) For the given {Φ0[n]}, {α} k,0 [n]}、{β j,0 [n]}, {q0[n]} and {v0[n]}, solve the convex optimization problem (14) to obtain the optimization result of the (t+1)th iteration, that is, the UAV transmission power {p c,t+1 [n],p r,t+1 [n]} is used as the initial value for the (t+1)th iteration;

[0162] 3) For a given {p} c,t+1 [n],p r,t+1 [n]}、{Φ0[n]}、{α k,0 [n]}、{β j,0 [n]}, solve the convex optimization problem (26) to obtain {q t+1 [n]} and {v t+1 [n]}, the number of update iterations is t = t + 1;

[0163] 4) For a given {p} c,t+1 [n],p r,t+1 [n]}、{q t+1 [n]}、{v t+1 [n]}、{α k,0 [n]}、{β j,0 [n]}, solve the convex optimization problem (34) to obtain {Φ t+1 [n]}, the number of update iterations is t = t + 1;

[0164] 5) For a given {p} c,t+1 [n],p r,t+1 [n]}、{q t+1 [n]}、{v t+1 [n]}、{Φ t+1 [n]}、{α k,0 [n]}、{β j,0 [n]}, solve the convex optimization problem (39), and thus obtain {α k,t+1 [n]} and {β j,t+1 [n]}, the number of update iterations is t = t + 1;

[0165] 6) Calculate the increase in average achievable rate after this iteration. If it is greater than the threshold ε0, update the iteration number t = t + 1 and jump to step 2) for the next iteration optimization. If the increase in the optimization target value is less than the threshold ε0, terminate the iteration and output the corresponding optimization variable value.

[0166] This embodiment verifies:

[0167] The horizontal coordinates of the reflector and the eavesdropper are p. i =[0,5] and p e = [6,6]m, the height of the reflector is 5m, the drone's flight altitude is 20m, and the horizontal coordinates of the initial and final points are q. I = [-60, 45]m and q F = [60,45]m. The number of users and targets is 2 each. The horizontal coordinates of the two users are u1 = [-6,10] and u2 = [6,10]m, respectively. The horizontal coordinates of the two targets are t1 = [-25,15] and t2 = [25,15]m, respectively. The path loss exponent κ from the UAV to the users and targets is 3.3, the path loss exponent α from the UAV to the reflector is 2.2, the channel gain ρ from the UAV to the reference point is -28dB, and the noise power σ... 2 = -80dBm. Other parameters related to UAV flight are set to flight time slot δ. t =1s, the maximum speed of the drone is v max =6m / s, minimum velocity v min =3m / s, maximum acceleration a max =3m / s 2 .

[0168] (1) The method proposed in this embodiment adopts continuous convex approximation, Riemannian manifold optimization and coordinate descent, and considers two benchmarks for comparison: one is that the UAV flies from the starting position to the ending position at the same speed; the other is that no intelligent reflective surface is used to assist in the integration of communication and perception.

[0169] (2) Analyze the drone trajectories under different optimization schemes:

[0170] Figure 2 Optimal UAV trajectories for the proposed scheme and benchmarks are presented. For both the proposed scheme and the reflectorless scheme, the UAV first flies from its initial position towards the user and target at maximum speed, then hovers over the optimal position. Finally, the UAV flies to its final position at maximum speed. The reason is that the UAV can obtain better channel gain when it is closer to the user and target, thus allowing more time to hover over them to support the integrated sensing service. Furthermore, since there is no eavesdropping on channel state information, the UAV will not fly away from the eavesdropper. In this scheme, the UAV tends to approach the reflector to obtain higher passive beamforming gain. Simultaneously, the distance between the optimal hovering position and the reflector decreases as M decreases.

[0171] (3) The performance of different schemes as a function of maximum transmit power is analyzed:

[0172] Figure 3 shows the performance curves of different schemes as a function of maximum transmit power. Clearly, the average achievable rate and secure rate of all schemes increase with P. max The signal-to-noise ratio (SNR) increases with increasing transmit power. On the one hand, the higher the transmit power, the greater the SNR at each user. On the other hand, using sensor signals to suppress eavesdropping can effectively improve security. Furthermore, the proposed scheme outperforms other benchmark schemes in both average achievable rate and secure rate. Therefore, jointly optimizing the trajectory and the phase shift of the reflector brings further performance gains. Moreover, due to the passive beamforming gain of the reflector, the straight-flying scheme outperforms the scheme without a reflector.

[0173] (4) The performance of different schemes is analyzed as the number of reflective surface units changes:

[0174] Figure 4 shows the performance curves of different schemes as a function of the number of reflector elements. Since the reflector introduces more passive beamforming gain, the average reachability and security of the reflector-assisted scheme both increase with the number of reflector elements. However, the performance of the reflectorless scheme remains unchanged with increasing reflector element count and is worse than the proposed scheme. Furthermore, the proposed scheme outperforms the straight-line flight scheme because it achieves higher channel gain by optimizing the UAV's trajectory. Moreover, the performance gap between the straight-line flight scheme and the reflectorless scheme decreases with increasing reflector element count; when the number of reflector elements is very large, the straight-line flight scheme outperforms the reflectorless scheme.

[0175] (5) Analyze the performance changes of different schemes with flight time:

[0176] Figure 5 shows the performance curves of different schemes as a function of flight time. For all schemes, the average reachability and security rate increase with increasing flight time. Compared with other schemes, the performance of straight-line flight increases much more slowly with flight time due to its high path loss. For the proposed scheme and the reflectorless scheme, as the flight time increases, the UAV can spend more time hovering over the user to fully utilize the high channel gain, thereby improving system performance. Through joint optimization of UAV trajectory and phase shift, this scheme can achieve better performance than the benchmark scheme. Furthermore, due to the passive beamforming gain provided by the reflector, the straight-line flight scheme outperforms the reflectorless scheme when the flight time is short. However, the performance gap decreases with flight time, and when the flight time is sufficiently long, the reflectorless scheme outperforms the straight-line flight scheme.

[0177] (6) Analyze the performance changes of different schemes with flight time:

[0178] Figure 6 The user scheduling results of the proposed scheme are presented. In the first 20 seconds, the drone communicates with user 1 and senses target 1; the remaining time it serves user 2 and senses target 2. This is because, during the first 20 seconds, the distance between the drone and user 1 is less than the distance between the drone and user 2. Therefore, the drone communicates with user 1 to maximize the channel gain of the drone-user 1 link. However, during the rest period, the distance between the drone and user 1 is greater than the distance between the drone and user 2. Therefore, the drone serves user 2 to improve performance.

[0179] The above-described embodiments are merely illustrative of the implementation methods of the present invention, but should not be construed as limiting the scope of the present invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these modifications and improvements all fall within the protection scope of the present invention.

Claims

1. A design method for integrated security communication and perception of unmanned aerial vehicles (UAVs) assisted by intelligent reflective surfaces, characterized in that, Includes the following steps: The first step is to build a system model; The second step is to determine the objective function and optimization variables, and to list the optimization problem. By analyzing the drone's trajectory {q[n]}, speed {v[n]}, and power allocation {p}... c [n]} and {p r [n]}, the phase shift matrix of the intelligent reflector {Φ[n]}, user scheduling {α k [n]} and target scheduling {β j [n]}; In this optimization problem, P max Indicates the maximum instantaneous transmit power. Γ represents the average transmit power. j [n] represents the power threshold of the j-th target in the n-th time slot, C1-C2 represent power constraints, C3-C4 represent serial interference cancellation constraints, C5 represents binary constraints, C9 represents received power constraints, C10-C11 represent trajectory constraints, and C13-C14 represent velocity constraints. The third step is to design an algorithm to solve the optimization problem: Using the idea of ​​block iteration, the optimization problem shown in formula (13) is decomposed into four sub-problems, and the power allocation problem is solved using CVX. For UAV trajectory optimization, the non-convex problem is approximately transformed into a convex optimization problem by using the continuous convex approximation method. For the reflector phase optimization problem, the Riemannian manifold optimization and penalty term method are used to solve it. For the target and user timing problem, the penalty function method is used to solve it.

2. The integrated design method for UAV security communication and perception assisted by an intelligent reflective surface as described in claim 1, characterized in that, The first step, constructing the system model, is as follows: (1) The UAV is used as a dual-function base station to communicate with K ground users and sense J ground targets. At the same time, there is a potential eavesdropper who eavesdrops on the users' information, and the channel state information of the eavesdropper cannot be obtained. The base station is a single antenna, and the users are single antennas. In the three-dimensional coordinate system, the horizontal coordinates of the k-th (k=1,...,K) user and the j-th (j=1,...,J) target are u k =(x u,k ,y u,k ) T and t j =(x t,j ,y t,j ) T The horizontal coordinate of the eavesdropper is p. e =(x e ,y e ) T ; (2) Assume the drone travels at a fixed altitude H u Flight, with a flight time of T, is discretized into N time slots, with time slot lengths of... Then the horizontal coordinate of the UAV in the nth time slot is q[n] = (q x [n],q y [n]) T From the initial position q I Fly to the destination position q F That is, q[0] = q I ,q[N]=q F ; (3) Assuming the channel between the base station and the user is a line-of-sight channel, then the channel between the UAV and user k, target j and the eavesdropper in the nth time slot is: Where L0 represents the channel gain per unit reference distance, H represents the distance between the UAV and ground user k in the nth time slot. u This represents the drone's altitude; α is the path loss index. Let t be the distance between the UAV and target j in the nth time slot. j This represents the horizontal coordinate of the j-th (j = 1, ..., J) target; Let p be the distance between the drone and the eavesdropper in the nth time slot. e The horizontal coordinates of the eavesdropper; The channel gain from the UAV to the smart reflector is expressed as: Where κ represents the path loss exponent. a(θ) represents the distance between the base station and the smart reflector. ai [n],φ ai [n]) represents the steering vector at time slot n, where θ ai [n] represents the pitch angle in the nth time slot, φ ai [n] represents the azimuth angle of the nth time slot, and the specific expression is shown in formula (3): in, This represents the phase delay of the steering vector along the x-axis, x i q represents the x-axis coordinate value of the intelligent reflective surface. x [n] represents the x-axis coordinate value of the UAV in the nth time slot, d ai [n] represents the distance between the UAV and the reflector in the nth time slot; The y-axis represents the phase delay of the steering vector; λ represents the wavelength; d represents the element spacing; M x M represents the number of units on the x-axis of the reflecting surface. y This indicates the number of units on the y-axis of the reflecting surface; j represents the imaginary unit. Indicates the Kronecker product; The channel from the smart reflector to the ground user is also a line-of-sight link, and its gain is expressed as: in, This indicates the distance between the smart reflective surface and the user. This indicates the phase delay of the steering vector along the x-axis. This indicates the phase delay of the steering vector along the y-axis. Denotes the Kronecker product, x u,k This represents the x-axis coordinate of the k-th user; The phase shift adjustment of the signal on the smart reflector can be expressed as: Where diag represents a diagonal matrix, θ m [n](θ∈[0,2π]) is the phase shift value of the m-th reflection element in the n-th time slot; (4) Assuming that the UAV uses time division multiple access to eliminate multi-user interference, that is, the UAV only communicates with one user in a specific time slot, the following constraints apply: If the drone serves user k in the nth time slot, α k [n] = 1; if the drone serves other users in the nth time slot, α k [n] = 0, where α k [n] represents the associated indicator variable for the k-th user in the n-th time slot; k = 1, ..., K represents the user index; (5) Assume that in order to reduce the computational complexity of parameter estimation, the UAV will detect at most one target in a specific time slot; in order to ensure perception performance, the j-th target should be detected at least N times. j Therefore, the following constraints apply: If the UAV detects target j in the nth time slot, β j [n] = 1; if the UAV detects other targets in the nth time slot, β j [n] = 0, β j [n] represents the associated indicator variable for the j-th target in the n-th time slot; j = 1, ..., J represents the target index; (6) The transmitted signal in the nth time slot is: in, This represents the communication symbol of the k-th user; Indicates a sensed signal; p c [n] and p r [n] represents the power allocated to the communication and sensing signals, respectively; K represents the number of users; J represents the number of targets; This is a complex Gaussian distribution with a mean of 0 and a variance of 1. (7) Assume that each user uses serial interference cancellation to eliminate interference from the sensed signal, that is, each user removes the sensed signal before decoding its own signal; in order to ensure serial interference at each user, the following constraints apply: in, and Let represent the equivalent channels from the UAV in the nth time slot to user k and target j, respectively; This represents the channel from the UAV in the nth time slot to user k; Φ[n] represents the channel from the reflector to user k; Φ[n] represents the phase shift matrix of the nth time slot reflector. This represents the channel from the UAV in the nth time slot to the reflector. This represents the channel from the UAV in the nth time slot to target j; This represents the channel from the UAV to the target in the nth time slot; (8) The reachable rate of user k in the nth time slot can be expressed as: Where, σ 2 Indicates noise power; α k [n] represents the associated indicator variable for the k-th user in the n-th time slot; p c [n] represents the power allocated to the communication signal in the nth time slot; h k [n] represents the equivalent channel from the nth time slot UAV to user k; (9) The received power of the j-th target in the n-th time slot can be expressed as: p c [n] and p r [n] represents the power allocated to the communication and sensing signals, respectively; This represents the equivalent channel from the UAV in the nth time slot to the target j; (10) Assuming the eavesdropper does not employ serial interference cancellation, the eavesdropping rate in the nth time slot can be expressed as: in, h represents the equivalent channel between the drone and the eavesdropper in the nth time slot; ae [n] represents the channel from the drone to the eavesdropper in the nth time slot; h ie This represents the channel from the nth time slot reflector to the eavesdropper; (11) The average safety rate can be expressed as: in,[·] + =max(·,0),R k [n] represents the communication rate of user k in the nth time slot; R e [n] represents the eavesdropping rate of the nth time slot.

3. The integrated design method for UAV security communication and perception assisted by an intelligent reflective surface as described in claim 1, characterized in that, The third step involves designing an algorithm to solve the optimization problem, as detailed below: (1) Optimize power allocation {p c [n],p r [n]} Given {Φ[n], α k [n],β j [n],p c [n],p r The optimization problem is represented as [n]}. This problem is a convex optimization problem, which can be solved using CVX; (2) Optimize the trajectory {q[n]} and velocity {v[n]} Given {Φ[n], α k [n],β j [n],p c [n],p r The optimization problem is represented as [n]}. The trust region continuous convex approximation is used to approximate h. ai The guide vector in [n]; To ensure approximate accuracy, the following constraints apply. Where δ represents the precision threshold; Official 15, in, This represents the distance vector between the UAV in the nth time slot and user k; and This represents the approximate channel between the UAV in the nth time slot and the reflector. Similarly, Represented as in, This represents the approximate channel between the UAV in the nth time slot and the target j. This represents the distance vector between the UAV in the nth time slot and the target j. Let p[n] represent a Hermitian matrix, where p[n] = p c [n]+p r [n] represents the transmit power of the UAV in the nth time slot; Therefore, the optimization problem is expressed as By introducing slack variable {γ k [n]}, which satisfies The optimization problem is equivalently represented as By introducing slack variables and Its satisfaction The optimization problem (22) is equivalently represented as Furthermore, by substituting q[n] into... d ai [n] and (25) can be equivalently expressed as These constraints are in the form of convexity; by continuous convex approximation, (26) can be transformed into a convex problem as shown in (28). stC1,C2,C3,C9 Formula (28) can then be solved using CVX; (3) Optimize the phase of the reflecting surface {Φ[n]}, Given {p c [n],p r [n],q[n],v[n],α k [n],β j The optimization problem for [n] can be represented as: definition p c [n]|h k [n]| 2 and It can be equivalently represented as in, Represents the augmented phase shift vector; Let represent the equivalent channel matrix from the UAV in the nth time slot to user k; Let represent the equivalent channel matrix from the UAV in the nth time slot to the target j; Therefore, the optimization problem can be equivalently represented as Furthermore, the above problems can be processed in parallel. Specifically, the phase optimization problem in the nth time slot can be expressed as: By moving constraints C1 and C2 from problem (30) to the objective function, the optimization problem can be expressed as follows: in, in, This represents the penalty coefficient; because of the constraints in (34) It is a complex circular manifold, and we can solve it using the manifold conjugate gradient algorithm; (4) Optimize {α k [n]} and {β j [n]} Given {p c [n],p r The optimization problem is represented as [n],q[n],v[n],Φ[n]}. C5 can be equivalently represented as They are transformed into the following convex form using continuous convex approximation: By adding constraint (38) as a penalty term to the objective function of (36), problem (36) can be rewritten as follows: Where γ>0 represents the penalty coefficient, this problem is a convex optimization problem, which can be solved using CVX; (5) The optimization problems shown in formulas 14, 28, 34 and 39 above are solved by alternating optimization algorithms. In each iteration, the power allocation problem is solved by CVX, then the UAV trajectory optimization is solved by continuous convex approximation algorithm, then the user and target scheduling is solved by continuous convex approximation and penalty function method, and then the reflector phase optimization problem is solved by Riemannian manifold optimization. Finally, the values ​​of the parameters are updated for the next iteration until the algorithm converges.

4. The intelligent reflective surface-assisted UAV security communication and sensing integrated design method according to claim 3, characterized in that, The third step (5) is as follows: 1) Set the initial power allocation {p c,0 [n],p r,0 [n]}, intelligent reflector reflection phase {Φ0[n]}, user scheduling {α k,0 [n]}, target scheduling {β j,0 [n]}, UAV flight trajectory {q0[n]}, UAV flight speed {v0[n]}, iteration count t=0; 2) Solve the convex optimization problem (14) to obtain the power allocation result of the t-th iteration, i.e., {p c,t [n],p r,t [n]}; 3) Solve the convex optimization problem (26) to obtain the trajectory and velocity optimization results for the t-th iteration, i.e., {q t [n],v t [n]}; 4) Solve the optimization problem (34) to obtain the optimization result of the reflector phase in the t-th iteration, i.e., {Φ t [n]}; 5) Solve the convex optimization problem (39) to obtain the optimization results of the user and target scheduling in the t-th iteration, i.e., {α k,t [n],β j,t [n]}; 6) Update t = t + 1; skip to step 2) for the next iteration optimization; 7) Until convergence; the convergence condition shown is: reaching the maximum number of convergences or if the increase in the optimization objective value is less than the threshold ε0.

Images