Optimization Method for Integrated Sensor System of Unmanned Aerial Vehicle with Intelligent Reflective Surface

Abstract

This application provides an optimization method for an integrated UAV sensing system with an intelligent reflective surface. The system includes a UAV, an intelligent reflective surface, a set of ground users, a set of targets, a base station, and a controller. The method steps are as follows: Based on the UAV trajectory, time allocation, and AIRS phase offset parameters, models for communication data collection and sensing data collection are constructed respectively; an objective function and constraints for system optimization are constructed, with the weighted sum of communication data collection and sensing data collection maximizing the objective function; the UAV trajectory, time allocation, and AIRS phase offset parameters are iteratively updated using an alternating optimization method; the objective function is calculated based on the updated parameters; the iteration is repeated until the objective function converges, and the optimized UAV trajectory, time allocation, and AIRS phase offset parameters are output. This application effectively balances communication and sensing performance, exhibiting stable performance gains.

Description

Technical Field

[0001] This invention relates to the field of mobile airborne intelligent reflective surfaces, and in particular to an optimization method for an integrated sensor system for unmanned aerial vehicles (UAVs) with intelligent reflective surfaces. Background Technology

[0002] The growing demand for ubiquitous connectivity in the 6G era has driven the development of non-terrestrial networks, in which drones play a crucial role. Their high mobility enables them to operate as aerial base stations or relay nodes, maintaining favorable line-of-sight links with ground users in remote areas, traffic hotspots, and emergency scenarios. While drones offer flexibility, their ability to actively manage the wireless environment is limited. Combining reconfigurable smart surfaces with drones (known as aerial RIS or AIRS) can enhance signal control capabilities.

[0003] Unlike fixed-ground reconfigurable smart surfaces (RIS), Airborne Reconfigurable Smart Surfaces (AIRS) dynamically reconfigure the propagation environment using the maneuverability of UAVs to improve link reliability and transmission efficiency. Previous research has explored Airborne Reconfigurable Smart Surfaces (AIRS) support systems from various perspectives. One study proposed a joint placement and passive beamforming problem aimed at achieving worst-case signal-to-noise ratio coverage. Another study employed Q-learning and alternating optimization to improve the downlink signal-to-noise ratio of high-altitude platforms under Mahalanobis jitter conditions. Yet another study derived analytical expressions for the outage probability and throughput. Still another study proposed a control framework based on Euler angles to jointly optimize UAV trajectory, altitude, and phase shift while considering beam misalignment. Beyond communications, emerging applications require airborne platforms to simultaneously support reliable data transmission and efficient sensing, which is crucial for tasks such as target monitoring and situational awareness.

[0004] Beyond high-capacity communication, future wireless systems are expected to support high-resolution environmental perception. One research proposed a self-sensing reconfigurable smart surface (RIS) for target localization. In this study, a smart reflector controller transmits a probe signal, and a sensor receives the echo signal, effectively mitigating the severe path loss problem of traditional passive sensing modes. Addressing the limitations of fixed deployment and environmental obstruction on ground-based implementations, extending self-sensing smart reflectors to UAV-borne platforms is a natural solution. Combining the sensing capabilities of the smart reflector with the mobility of the UAV can simultaneously support reliable communication and high resolution. This dual-function AIRS integrates signal transmission and sensing capabilities, thereby enhancing adaptability in dynamic environments. However, challenges remain, including the trade-off between time resource allocation for sensing and communication, spatial trajectory design coupled with resource scheduling, and sensitivity to time-varying angles affecting effective array gain. Summary of the Invention

[0005] The purpose of this invention is to provide an optimization method for an integrated sensing and communication system for unmanned aerial vehicles (UAVs) with an intelligent reflective surface. This method addresses the technical challenges of simultaneously addressing the issues of time resource allocation, resource scheduling coupling, and spatial trajectory design between sensing and communication in an integrated sensing and communication system for UAVs.

[0006] This application provides an optimization method for an integrated sensing and communication system for unmanned aerial vehicles (UAVs) with an intelligent reflective surface. The system includes a UAV, an intelligent reflective surface, a set of ground users, a set of targets, a base station, and a controller. While sensing targets, the UAV uses the intelligent reflective surface to reflect signals between ground users and the base station. The controller controls the UAV's flight trajectory, the phase shift of the RIS (Range Identifier), and the time allocation between communication and sensing. The specific steps are as follows:

[0007] S1: Based on the system's UAV trajectory, time allocation, and AIRS phase offset parameters, construct models for the amount of communication data collected and the amount of perception data collected, respectively;

[0008] S2: Construct the objective function and constraints for system optimization, with the system that maximizes the weighted sum of communication data collection and sensing data collection as the objective function for optimization.

[0009] S3: Iteratively update the UAV trajectory, time allocation, and AIRS phase offset parameters using an alternating optimization method; and calculate the objective function based on the updated parameters;

[0010] S4: Repeat step S3 until the objective function tends to converge, and output the optimized UAV trajectory, time allocation and AIRS phase offset parameters.

[0011] Optionally, in step S1, a communication data collection volume model is constructed. for:

[0012] ;

[0013] in, The nth time slot is allocated to the user The communication transmission time, where B is the bandwidth. For transmission power, For the variance of additive white Gaussian noise, the user Equivalent concatenated channels are , The AIRS phase offset in the nth time slot; For users Channel to the smart reflector , , , As the guide vector, , , , The angle of arrival is determined by... as well as Calculations show that For the channel from the smart reflector to the base station, , , For the drone's trajectory, .

[0014] Optionally, in step S1, a model for the amount of sensing data collected is constructed. for:

[0015] ;

[0016] in, The target to be sensed by the intelligent reflector in the nth time slot. Time, For signal bandwidth, For transmission power, For radar pulse duration, The power spectral density (PSD) coefficient. To predict the variance of the error, Radar duty cycle; This is a round-trip sensing channel model. , Characterizing small-scale fading, Characterizing large-scale path loss To receive the array response, , It is the azimuth angle. This refers to the scale of the sensor assembly.

[0017] Optionally, the objective function constructed in step S2 is:

[0018]

[0019] in, For the weighted sum utility function used to balance communication and sensing, ;

[0020] The constraints on the objective function are:

[0021]

[0022]

[0023]

[0024]

[0025]

[0026]

[0027]

[0028] in, For each time slot duration, Let T be the total duration and N be the number of time slots; For ground users, For the target set, As a minimum communication data volume constraint, To minimize the amount of data to be perceived, Let n be the propulsion power at time slot n. For airborne energy budget, For the phase of the f-th unit, the trajectory begins and These are the starting and ending points of the drone's trajectory, respectively.

[0029] Optionally, the specific steps in step S3 are as follows:

[0030] S3.1: Set the UAV trajectory, time allocation, and AIRS phase offset parameters for the initial iteration, and the initial iteration index. ;

[0031] S3.2: UAV trajectory obtained based on the i-th iteration and time allocation Update the AIRS phase offset in the (i+1)th iteration. ;

[0032] S3.3: Drone trajectory based on the i-th iteration AIRS phase offset with the (i+1)th iteration Update the time allocation for the (i+1)th iteration. ;

[0033] S3.4: AIRS phase shift based on the (i+1)th iteration and time allocation Update the drone trajectory in the (i+1)th iteration. ;

[0034] S3.5: Let i = i + 1, based on the drone trajectory of the (i + 1)th iteration Time allocation and AIRS phase offset Calculate the objective function for the (i+1)th iteration. .

[0035] Optionally, in step S3.2, the AIRS phase offset of the (i+1)th iteration is updated. The method is as follows: using the UAV trajectory obtained in the i-th iteration and time allocation With a fixed value, and aiming to maximize the total communication data collection volume of K users in N time slots, solve for the AIRS phase offset. .

[0036] Optionally, in step S3.3, the time allocation for updating the (i+1)th iteration is updated. The method is as follows: using the drone trajectory of the i-th iteration AIRS phase offset with the (i+1)th iteration Given a fixed value, with the objective function of system optimization as the goal, the time allocation is solved. .

[0037] Optionally, in step S3.4, the drone trajectory of the (i+1)th iteration is updated. The method is as follows: using the AIRS phase offset of the (i+1)th iteration and time allocation With a fixed value, and aiming to maximize the objective function of system optimization, the CVX method is used to solve for the time allocation. .

[0038] Optionally, the specific method in step S4 is as follows:

[0039] judge Whether it is true or not, in which: The convergence threshold, ;

[0040] If true, then the objective function The trajectory tends to converge, based on the drone trajectory in the (i+1)th iteration. Time allocation and AIRS phase offset Output as the optimal parameters;

[0041] If this condition is not met, then the drone trajectory of the (i+1)th iteration will be... Time allocation and AIRS phase offset Return to step S3.2 and perform the (i+2)th iteration.

[0042] Because of the adoption of the above technical solution, the present invention has the following advantages:

[0043] This application effectively balances communication and sensing performance by jointly optimizing the UAV's flight trajectory, RIS phase shift, and communication and sensing time allocation, and reveals a threshold-based time allocation mechanism with stable performance gains.

[0044] Other advantages, objectives, and features of the invention will be set forth in part in the description which follows, and in part will be apparent to those skilled in the art from the following examination, or may be learned from practice of the invention. The objectives and other advantages of the invention can be realized and obtained through the following description. Attached Figure Description

[0045] The accompanying drawings of this invention are described below.

[0046] Figure 1 This is a flowchart of the optimization method for the UAV integrated sensing system of the present invention.

[0047] Figure 2 This is the optimized drone trajectory diagram used in the simulation of this invention.

[0048] Figure 3 This is the optimized time allocation diagram used in the simulation of this invention.

[0049] Figure 4 This is the optimal phase shift distribution diagram in the simulation of this invention when the drone flies away from the user.

[0050] Figure 5 This is the optimal phase shift distribution diagram when the drone is directly above the user in the simulation of this invention.

[0051] Figure 6 This is a graph showing the change in the total system utility as a function of the number of intelligent reflective surface units in the simulation of this invention.

[0052] Figure 7 This is a graph showing the variation of the stochastic energy budget of the total system utility in the simulation of this invention. Detailed Implementation

[0053] The present invention will be further described below with reference to the accompanying drawings and embodiments. In the description of the embodiments of the present invention, it should be noted that, unless otherwise explicitly specified and limited, the terms "connected" or "linked" should be interpreted broadly. For example, it can be a fixed connection, a detachable connection, an integral connection, an electrical connection, or a signal connection; it can be a direct connection or an indirect connection through an intermediate medium.

[0054] Example 1:

[0055] A drone-based integrated sensing system with an intelligent reflective surface includes a drone, an intelligent reflective surface (RIS), and a ground user (GU) cluster. Target set Base stations (BS) and controllers;

[0056] The intelligent reflective surface (RIS) is mounted on the drone. The intelligent reflective surface (RIS) is a uniform planar array with... A reflective unit; the UAV can use a smart reflective surface (RIS) to reflect signals between ground users (GU) and base stations (BS) while sensing targets, bypassing obstacles such as buildings and trees that block direct line-of-sight communication; the controller is used to control the flight trajectory of the UAV, the phase offset of the RIS, and the time allocation of communication and sensing.

[0057] The set of ground users (GUs) is as follows ,in: , The target set is ,in: , Base station (BS) is located in During the communication phase, the controller integrated into the smart reflector transmits sensing signals and uses a receiving sensor integrated on the smart reflector to capture the reflected echo. During the sensing phase, the reflective units of the passive smart reflector (RIS) are temporarily turned off, thereby isolating the active sensor and avoiding self-interference.

[0058] In this embodiment, the controller employs a Time Division Multiple Access (TDMA) protocol, dividing the total duration T into N time slots, each time slot lasting for a duration of [duration missing]. In each time slot n, the total duration is The allocation is between K ground users and M sensing targets, let Indicates user allocation The time of communication transmission makes This indicates that the intelligent reflective surface controller senses the target. The allocation is subject to time constraints. .

[0059] Then: the trajectory of the UAV in the nth time slot is The altitude H is fixed to reduce energy consumption, and the speed of the drone is limited by... That is, satisfying The trajectory begins finally ,Right now and The time allocation for communication and sensing in the nth time slot is as follows: = Phase shift of the RIS in the nth time slot Where: f represents the phase of the f-th unit, .

[0060] Example 2:

[0061] An optimization method for an integrated sensor system for unmanned aerial vehicles (UAVs) with an intelligent reflective surface, comprising the following steps:

[0062] S1: Based on the system's UAV trajectory, time allocation, and AIRS phase offset parameters, construct models for communication data collection volume and perception data collection volume, respectively; the specific steps are as follows:

[0063] S1.1: Constructing a communication data collection volume model: During the communication phase, the UAV assists ground users in uplink transmission to the base station via an aerial intelligent reflector. The air-to-ground channel is dominated by the line-of-sight (LoS) component. Constructing a communication data collection volume model. for:

[0064] (1)

[0065] in, The nth time slot is allocated to the user The communication transmission time, where B is the bandwidth. For transmission power, For the variance of additive white Gaussian noise, the user Equivalent concatenated channels are , The AIRS phase offset in the nth time slot; For users Channel to the smart reflector , , , As the guide vector, , , , The angle of arrival is determined by... as well as Calculations show that For the channel from the smart reflector to the base station, , , For the drone's trajectory, .

[0066] In this embodiment, to meet the uplink transmission requirements, each user must meet the following conditions: ;in: This is a minimum communication data volume constraint.

[0067] S1.2: Constructing a model for the amount of sensory data collected: for the target Active sensing is the interval allocated in each time slot n. The process involves: an internal intelligent reflector (RIS) controller transmitting a detection signal to the target; this signal, after being reflected by the target, being received by airborne sensors linearly arranged on the RIS; processing the signal to extract relevant target information; and a sensor configuration of [scale missing]. A uniform linear array, wherein the guiding vector is... , For normalized phase shift; the receiver array response is ,in: It is the azimuth angle. The sensor spacing; the round-trip sensing channel is modeled as ,in: Characterizing small-scale fading, Characterize large-scale path loss. Construct a model of the amount of sensing data collected. for:

[0068] (2)

[0069] in, For signal bandwidth, For transmission power, For radar pulse duration, The power spectral density (PSD) coefficient. To predict the variance of the error, This refers to the radar duty cycle; to ensure adequate monitoring, the accumulated sensing data must meet certain requirements. ,in: This is a constraint based on the minimum amount of data to be perceived.

[0070] In this embodiment, the UAV's energy consumption mainly consists of three parts: propulsion energy consumption, communication energy consumption, and sensing energy consumption. In real-world scenarios, propulsion energy consumption is primarily in the kilojoule range, while communication and sensing energy consumption are only in the joule range. Therefore, the focus is on propulsion energy, while other components are ignored; the propulsion power at time slot n is... ,in: and The blade drag power and induced power in hovering state. and The tip velocity and the mean induced velocity are given. These are the airframe drag ratio, air density, rotor solidity, and rotor disk area. The total propulsion energy consumption during the mission cycle is... This ensures that the operation stays within the onboard energy budget.

[0071] S2: Construct the objective function and constraints for system optimization, using the system with the weighted sum of the collected communication data and the collected sensing data as the objective function; the specific method is as follows:

[0072] System performance depends on the UAV trajectory, time allocation, and reconfigurable smart reflector phase shift. The trajectory influences the communication channel and sensing geometry, while time allocation controls resource allocation between communication transmission and target perception. The smart reflector phase shift affects signal propagation for both functions. Communication performance is measured by the total amount of communication data collected, while perception performance is measured by the cumulative amount of sensing data collected. Since enhanced perception typically requires more time and a more favorable geometry, thus reducing communication opportunities, these two objectives are conflicting. To find a balance between these two conflicts, a parameterized weighted sum utility function is employed. To balance communication and sensing. This joint optimization problem ( The aim is to maximize weighted utility under constraints of communication, perception, energy, and mobility by optimizing UAV trajectory, time allocation, and reconfigurable intelligent reflector phase shift. The constructed objective function is:

[0073]

[0074] in, For the weighted sum utility function used to balance communication and sensing, ;

[0075] The constraints on the objective function are:

[0076]

[0077]

[0078]

[0079]

[0080]

[0081]

[0082]

[0083] question( Because there is a strong coupling between the UAV trajectory, time allocation and intelligent reflector configuration, and the objective function and constraints are non-convex.

[0084] S3: Iteratively update the UAV trajectory, time allocation, and AIRS phase offset parameters using an alternating optimization method, and calculate the objective function based on the updated parameters; the specific steps are as follows:

[0085] S3.1: Set the drone trajectory for the initial iteration Time allocation and AIRS phase offset Parameters, initial iteration index ;

[0086] S3.2: UAV trajectory obtained based on the i-th iteration and time allocation Update the AIRS phase offset in the (i+1)th iteration. Specifically:

[0087] The drone trajectory obtained in the i-th iteration and time allocation For a fixed value, the AIRS phase offset in the (i+1)th iteration Optimization is needed; due to the amount of target perception data collected. Primarily determined by the direct backscatter link, it is unaffected by the intelligent reflector phase configuration designed for communication; therefore, the problem degenerates into maximizing the total uplink data volume, expressed as:

[0088]

[0089]

[0090] It can be verified by contradiction that the maximization problem () The objective function in () is equivalent to maximizing the scheduled user C in each time slot n. k Effective cascaded channel gain Effective scalar channel Represented as ,in: Let be the complex channel coefficient of the f-th reflection unit. To derive the optimal phase shift, the above complex coefficients are expressed in polar coordinates as follows: , Substitute the above parameters into Then, the composite channel gain can be expressed as = According to the triangle inequality for complex numbers, The upper bound can be derived as follows: ;in: Let f represent the composite phase of the f-th propagation path. This upper bound is reached if and only if the phases of all F terms in the summation are aligned to the same constant C. Without loss of generality, let C = 0. By setting the phase of each term to zero, we can obtain After rearranging and transforming each term, the optimal phase-shift closed-form solution for the f-th reflecting element can be derived, namely:

[0091]

[0092] It should be noted that, although the closed-form solution above may exceed... Range, but phase response The periodicity ensures that the optimal phase shift can be achieved through the modulus. The operation uniquely maps to the feasible region without sacrificing optimality.

[0093] S3.3: Drone trajectory based on the i-th iteration AIRS phase offset with the (i+1)th iteration Update the time allocation for the (i+1)th iteration. Specifically:

[0094] The drone trajectory in the i-th iteration AIRS phase offset with the (i+1)th iteration For a fixed value, allocate time for the (i+1)th iteration. Optimize; described as:

[0095]

[0096]

[0097] By introducing the Lagrange multipliers corresponding to the constraint conditions (4)–(6) The Lagrange function can be constructed. This function simultaneously represents both system utility and penalty term for constraint violation. To improve numerical stability and avoid allocation oscillations between iterations, a quadratic regularization term is introduced. This allows for a smoother time allocation. By recombining the terms for each time slot and ignoring the constant term, the Lagrange function can be expressed as: , - ,in, These represent the effective marginal utility of communication and sensing, respectively. These physical quantities can be understood as weighted gains in the time allocation of communication and sensing, where the dual variables dynamically adjust service priorities to meet long-term performance requirements.

[0098] The optimal solution can be derived from the Caro-Kuhn-Tucker (KKT) conditions, thus yielding the first-order optimality equation. and Solving the above equation yields a closed-form solution. Considering the box constraint... The optimal time allocation is:

[0099]

[0100]

[0101] in, To determine the dual variable, a two-level update mechanism is adopted. In the outer multiplier... and The update is performed using the subgradient method, expressed as follows: , ,in, To decrease the step size. In the inner layer, for each time slot, for a given... and Since the first-order optimality condition is , and dual variables Since it is a monotonic function, its optimal value can be efficiently found through binary search, thus determining the optimal value. .

[0102] S3.4: AIRS phase shift based on the (i+1)th iteration and time allocation Update the drone trajectory in the (i+1)th iteration. Specifically:

[0103] The AIRS phase offset of the (i+1)th iteration and time allocation For a fixed value, solve the following subproblems ( To update the trajectory at the (i+1)th iteration, that is:

[0104]

[0105]

[0106] Due to the coupled dependence of channel gain and propulsion power on the UAV trajectory, this problem ( The complex channel coefficients are nonconvex. Specifically, the cascaded channel gain involves a nonlinear steering vector, and the propulsion model introduces additional nonconvexity through the UAV's velocity. To address this issue, a successive convex approximation framework is employed, using the solution obtained from the previous iteration to iteratively optimize the trajectory. To simplify the problem, the angular components are approximated using the UAV trajectory from the previous iteration. Specifically, in the (i+1)th iteration, the complex channel coefficients are approximated as... Therefore, users The amount of communication data is represented as ,in, This is treated as a constant in the current iteration. Similarly, based on the geometric configuration of the i-th iteration, the array response vector used for sensing is approximately... And aggregate the constant terms into coefficients. And the amount of perceived data can be expressed as To convex the above expression for the amount of data, slack variables are introduced. and To define the squared distance, that is:

[0107]

[0108]

[0109]

[0110] user The amount of communication data satisfies the following functional form: Since f(U, V) is convex with respect to U and V respectively, but the data requirement is a lower bound, a first-order Taylor expansion can be used at local points. Find the concave lower boundary. ,Right now: + , where gradient coefficients and By taking the partial derivatives, we obtain the following: , Similarly, at local points A concave lower bound is obtained at this point, namely: , where gradient coefficients The derivation is as follows: Propulsion energy The trajectory is non-convex with respect to the drone's trajectory. To address this issue, a slack variable is introduced. This is used to decouple nonlinear terms. Specifically, the induced power term can be constrained by an upper bound. It can be equivalently rewritten as The left-hand side is approximated by its first-order expansion at the local point obtained in the previous iteration. In the (i+1)th iteration, the linearized form is given by the following equation:

[0111]

[0112] in, , .

[0113] After this transformation, the original non-convex energy constraint can be approximated as:

[0114]

[0115] in, To reduce power loss.

[0116] Therefore, in the (i+1)th iteration, we update the trajectory by solving the following problem:

[0117]

[0118]

[0119] question( This is a standard convex optimization problem that can be solved efficiently using CVX.

[0120] S3.5: Let i = i + 1, based on the drone trajectory of the (i + 1)th iteration Time allocation and AIRS phase offset Calculate the objective function for the (i+1)th iteration. .

[0121] S4: Repeat step S3 until the objective function converges, and output the optimized UAV trajectory, time allocation, and AIRS phase offset parameters; the specific method is as follows:

[0122] judge Whether it is true or not, in which: The convergence threshold, ;

[0123] If true, then the objective function The trajectory tends to converge, based on the drone trajectory in the (i+1)th iteration. Time allocation and AIRS phase offset Output as the optimal parameters;

[0124] If this condition is not met, then the drone trajectory of the (i+1)th iteration will be... Time allocation and AIRS phase offset Return to step S3.2 and perform the (i+2)th iteration.

[0125] S5: Simulation and Verification

[0126] To evaluate the effectiveness of the proposed joint optimization method, this application constructs an aerial service simulation scenario with obstacles. A UAV equipped with an Airborne Intelligent Reflector (AIRS) simultaneously provides uplink communication and target perception services to ground users. The system includes K=3 ground users and M=2 perceived targets, randomly distributed within the coverage area. The UAV flies from a designated starting point to a designated destination, and the trajectory is jointly optimized during flight to maximize the utility of the sensor-weighted system. The simulation parameters are set as follows: the number of Airborne Intelligent Reflector elements is... Communication bandwidth B = 1 MHz, noise power =-100 dBm, sensing transmit power =0.1 W, user transmit power =0.1 W, total task duration T=30 s, minimum communication data volume constraint =0.1 Mbits, minimum perceived data volume constraint =10 Mbits.

[0127] Figures 1 and 2 show the optimized UAV trajectory and time allocation. As shown in Figure 1, the UAV trajectory varies with the weighting factor. Adaptive adjustment: When prioritizing perceived performance ( When the signal strength is 0.1, the drone flies closer to the target to improve the echo signal power; when communication performance is emphasized ( When the bandwidth is 0.9, the UAV flies closer to the ground user to obtain better channel conditions. As shown in Figure 2, the time allocation exhibits a proximity-driven scheduling characteristic, with the UAV allocating more transmission time to closer nodes, achieving coordinated optimization of mobility and resource management.

[0128] Figures 3 and 4 show The optimal phase shift distribution when ω = 0.9. (See diagram below.) Figure 4 As shown, in the 17th time slot, the drone is directly above the user, and the phase shift exhibits a diagonal fringe distribution, allowing it to simultaneously direct the signal beam towards the base station in both the horizontal and vertical dimensions; as Figure 3 As shown, in the 27th time slot, the drone flies away from the user, and the phase shift changes to a vertical stripe distribution. The phase gradient can be dynamically adjusted according to the real-time geometric position to suppress time-varying path differences, maintain coherent signal superposition, and maximize beamforming gain.

[0129] For comparative analysis, three representative benchmark schemes were selected: the Stochastic Perturbation Trajectory (SPT) scheme, the Potential Field Guided Trajectory (PFGT) scheme, and the Airborne Intelligent Reflector based on Eulerian Dynamics (ARIS-Euler) scheme. Figure 5 , 6 This presents a comparison of the total system utility under different conditions. Under the equilibrium weighting condition of 0.5, by Figure 5 As can be seen, the system efficiency increases significantly with the increase of the number of intelligent reflector units F, and the proposed scheme consistently outperforms all benchmark schemes. The ARIS-Euler scheme is limited by angle-dependent array gain loss and geometric mismatch; fixed-track and random-track schemes cannot achieve efficient coherent beamforming, making it difficult to fully utilize beamforming gain. Figure 6 It can be seen that the random energy budget The proposed solution can continuously optimize the UAV path to reduce path loss and maintain a better balance between communication data collection and perception performance. In contrast, the performance improvement of fixed trajectory and random trajectory solutions is limited, and the ARIS-Euler solution is affected by angle inaccuracy, which limits beamforming gain.

[0130] Experimental results show that the proposed self-sensing aerial intelligent reflector optimization method for integrated sensing can effectively coordinate UAV trajectory, time allocation and intelligent reflector phase shift design. It significantly outperforms existing benchmark schemes in terms of sensing performance trade-off and system utility improvement. It has stable performance gains under different intelligent reflector sizes, energy budgets and weighting coefficients, and has high efficiency and practicality.

[0131] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. An optimization method for an integrated sensing and communication system for unmanned aerial vehicles (UAVs) with an intelligent reflective surface, the system comprising a UAV, an intelligent reflective surface, a set of ground users, a set of targets, a base station, and a controller; the UAV uses the intelligent reflective surface to reflect signals between ground users and the base station while sensing targets, and the controller is used to control the UAV's flight trajectory, the phase shift of the RIS (Range Identification System), and the time allocation between communication and sensing; characterized in that, The specific steps are as follows: S1: Based on the system's UAV trajectory, time allocation, and AIRS phase offset parameters, construct models for the amount of communication data collected and the amount of perception data collected, respectively; S2: Construct the objective function and constraints for system optimization, with the weighted sum of communication data collection volume and sensing data collection volume as the objective function for system optimization; S3: Iteratively update the UAV trajectory, time allocation, and AIRS phase offset parameters using an alternating optimization method, and calculate the objective function based on the updated parameters; S4: Repeat step S3 until the objective function tends to converge, and output the optimized UAV trajectory, time allocation and AIRS phase offset parameters.

2. The optimization method for an integrated sensor system for unmanned aerial vehicles with an intelligent reflective surface according to claim 1, characterized in that, In step S1, a communication data volume model is constructed. for: ； in, The time slot allocated to the user in the nth time slot The communication transmission time, where B is the bandwidth. For transmission power, For the variance of additive white Gaussian noise, the user Equivalent concatenated channels are , The AIRS phase offset in the nth time slot; For users Channel to the smart reflector , , , As the guide vector, , , , The angle of arrival is determined by... as well as Calculations show that For the channel from the smart reflector to the base station, , , For the drone's trajectory, .

3. The optimization method for an integrated sensor system for unmanned aerial vehicles with an intelligent reflective surface according to claim 1, characterized in that, In step S1, a model for the amount of sensory data collected is constructed. for: ； in, The target to be sensed by the intelligent reflector in the nth time slot. Time, For signal bandwidth, For transmission power, For radar pulse duration, The power spectral density (PSD) coefficient. To predict the variance of the error, Radar duty cycle; This is a round-trip sensing channel model. , Characterizing small-scale fading, Characterizing large-scale path loss, To receive the array response, , It is the azimuth angle. This refers to the scale of the sensor assembly.

4. The optimization method for an integrated sensor system for unmanned aerial vehicles with an intelligent reflective surface according to claim 1, characterized in that, The objective function constructed in step S2 is: in, For the weighted sum utility function used to balance communication and sensing, ; The constraints on the objective function are: in, For each time slot duration, Let T be the total duration and N be the number of time slots; For ground users, For the target set, As a minimum communication data volume constraint, To minimize the amount of data to be perceived, Let n be the propulsion power at time slot n. For airborne energy budget, For the phase of the f-th unit, the trajectory begins and These are the starting and ending points of the drone's trajectory, respectively.

5. The optimization method for an integrated sensor system for unmanned aerial vehicles with an intelligent reflective surface according to claim 1, characterized in that, The specific steps in step S3 are as follows: S3.1: Set the UAV trajectory, time allocation, and AIRS phase offset parameters for the initial iteration, and the initial iteration index. ; S3.2: UAV trajectory obtained based on the i-th iteration and time allocation Update the AIRS phase offset in the (i+1)th iteration. ; S3.3: Drone trajectory based on the i-th iteration AIRS phase offset with the (i+1)th iteration Update the time allocation for the (i+1)th iteration. ; S3.4: AIRS phase shift based on the (i+1)th iteration and time allocation Update the drone trajectory in the (i+1)th iteration. ; S3.5: Let i = i + 1, based on the drone trajectory of the (i + 1)th iteration Time allocation and AIRS phase offset Calculate the objective function for the (i+1)th iteration. .

6. The optimization method for an integrated sensor system for unmanned aerial vehicles with an intelligent reflective surface according to claim 5, characterized in that, In step S3.2, update the AIRS phase offset for the (i+1)th iteration. The method is as follows: using the drone trajectory obtained in the i-th iteration and time allocation With a fixed value, and aiming to maximize the total communication data collection volume of K users in N time slots, solve for the AIRS phase offset. .

7. The optimization method for an integrated sensor system for unmanned aerial vehicles with an intelligent reflective surface according to claim 5, characterized in that, In step S3.3, update the time allocation for the (i+1)th iteration. The method is as follows: using the drone trajectory of the i-th iteration AIRS phase offset with the (i+1)th iteration Given a fixed value, with the objective function of system optimization as the goal, the time allocation is solved. .

8. The optimization method for an integrated sensor system for unmanned aerial vehicles with an intelligent reflective surface according to claim 5, characterized in that, In step S3.4, update the drone trajectory for the (i+1)th iteration. The method is as follows: using the AIRS phase offset of the (i+1)th iteration and time allocation With a fixed value, and aiming to maximize the objective function of system optimization, the CVX method is used to solve for the time allocation. .

9. The optimization method for an integrated sensor system for unmanned aerial vehicles with an intelligent reflective surface according to claim 5, characterized in that, The specific method in step S4 is as follows: judge Whether it is true or not, in which: The convergence threshold, ; If true, then the objective function The trajectory tends to converge, based on the drone trajectory in the (i+1)th iteration. Time allocation and AIRS phase offset Output as the optimal parameters; If this condition is not met, then the drone trajectory of the (i+1)th iteration will be... Time allocation and AIRS phase offset Return to step S3.2 and perform the (i+2)th iteration.

Images