Unmanned aerial vehicle assisted integrated optical access and backhaul resource allocation method oriented to user fairness

Abstract

The application claims a UAV-assisted optical communication and sensing integrated resource allocation method for user fairness, and relates to the field of optical communication networks, comprising: constructing a UAV-assisted O-ISAC system model, and further constructing a sensing model and a communication model based on the UAV-assisted ISAC system transmission model. A system optimization model is established to maximize the minimum average reachable rate in all ground terminals. To solve the proposed optimization problem, an alternating optimization algorithm based on BCD is designed, and P-SCA, a closed-form solution for sensing time and SCA method are combined for solving. The application takes the minimum average reachable rate as an index, and the effectiveness of the method is verified in the simulation results, and the superior multi-user fairness guarantee capability and the adaptability to the dynamic flight scene of the UAV are shown.

Description

Technical Field

[0001] This invention relates to the field of optical communication networks, and specifically to a user-fair, unmanned aerial vehicle (UAV)-assisted integrated optical and sensor resource allocation method. Background Technology

[0002] 6G networks place higher demands on communication speed, latency, and reliability. While traditional radio frequency (RF) communication offers advantages such as wide coverage and strong penetration, its spectrum resources are increasingly scarce, and its anti-interference performance is limited, making it difficult to meet diverse needs in complex environments. Free-space optical (FSO) utilizes unlicensed spectrum and short wavelengths in the optical band for information transmission, enabling high-speed communication and high-precision sensing. However, FSO communication relies on line-of-sight (LoS) and is susceptible to weather and obstacles. Unmanned aerial vehicles (UAVs), with their advantages of flexible deployment, cost-effectiveness, and ease of establishing LoS links, have become an important technology for expanding wireless network coverage and capacity. UAVs, by carrying onboard communication nodes, can effectively ensure that the LoS link between the transmitter and receiver is not blocked. However, the dynamic nature of UAVs further exacerbates the already stringent beam alignment issues in FSO communication.

[0003] Integrated Sensing and Communication (ISAC) deeply integrates radar detection and communication functions into a single system by sharing spectrum, hardware, and signal processing platforms, thereby greatly improving resource utilization efficiency and reducing system cost and power consumption. Compared with the rapidly developing radio frequency (RF) ISAC, the emerging Optical Integrated Sensing and Communication (O-ISAC) technology offers higher communication rates, stronger sensing accuracy, and lower interference. However, existing O-ISAC work mainly focuses on designing specific signal waveforms for the system, with sensing and communication largely confined to the level of service coexistence, without fully exploring functional mutual assistance. For UAV-assisted O-ISAC systems, although some research has begun to focus on sensing-assisted communication mechanisms, it typically establishes pointing error models under the assumption of UAV hovering or fixed position, making it difficult to accurately characterize the relationship between sensing angle error, beam deflection, and receiver radial offset under dynamic trajectories. Furthermore, UAV-assisted O-ISAC research focuses more on improving the overall system transmission rate, with insufficient consideration for multi-user service fairness, which can easily make it difficult for devices with poor link conditions or remote locations to obtain effective service.

[0004] Therefore, how to coordinate sensing accuracy, communication transmission performance, and multi-user fairness in UAV dynamic flight scenarios, and further improve the resource allocation efficiency of UAV-assisted O-ISAC systems, remains an urgent problem to be solved in current research on integrated optical and sensor resource optimization.

[0005] A search revealed application publication number CN120812715A, which discloses a method for trajectory planning and resource allocation in a drone-assisted sensor-computer fusion system. This method includes: S1: modeling the drone-assisted sensor-computer fusion system scenario; S2: modeling the channel model; S3: modeling the communication rate; S4: modeling the communication user data queue and the sensing data queue model; S5: modeling the drone trajectory uncertainty model; S6: modeling the system energy consumption model; S7: modeling the probability of false detection of eavesdroppers; S8: modeling system constraints; and S9: determining the drone trajectory planning and resource allocation strategy based on minimizing system energy consumption.

[0006] This patent is geared towards radio frequency communication-inductive-computing fusion systems, but does not address the impact of sensing errors on FSO beam pointing errors during UAV dynamic flight. Furthermore, its sensing function is primarily used to acquire sensing data and upload it to the base station for computational processing; sensing and communication still exist as parallel services, and no sensing-assisted communication mechanism is involved. In addition, while this solution sets a minimum user rate constraint, it is difficult to directly address the issue of fairness in multi-user services, as its primary goal is to minimize energy consumption.

[0007] To address the aforementioned issues, this invention, targeting the dynamic flight scenario of UAVs, establishes the relationship between sensing angle error, beam deflection, and receiver radial offset, and incorporates this relationship into the FSO channel model. Simultaneously, by jointly optimizing UAV trajectories, ground terminal association scheduling, and sensing time allocation, it maximizes the minimum average reachability among all terminals, thereby improving the fairness of multi-user services. Summary of the Invention

[0008] This invention aims to solve the problems of the prior art mentioned above. It proposes a drone-assisted integrated optical and sensor resource allocation method oriented towards user fairness. The technical solution of this invention is as follows:

[0009] A user-fair-oriented UAV-assisted optical-sensing integrated resource allocation method is proposed. This method addresses scenarios such as post-disaster emergency communication, temporary network coverage, and high-speed data transmission from sparsely populated ground terminals. By jointly designing a sensing-assisted beam alignment mechanism with UAV trajectory planning, ground terminal association scheduling, and sensing time allocation, it improves communication fairness among multiple ground terminals and enhances system resource utilization efficiency.

[0010] This invention considers a post-disaster emergency communication scenario, with the target area containing one UAV and K ground terminals (GTs). The UAV, acting as an airborne base station with both sensing and communication functions, provides sensing-assisted downlink data transmission services to multiple ground terminals during its mission cycle. The mission cycle is divided into multiple time slots, each further subdivided into sensing sub-slots and communication sub-slots. Within each time slot, the UAV first acquires the azimuth and elevation angle information of the served ground terminals through the sensing sub-slot, and then adjusts the FSO beam direction and performs downlink data transmission based on the sensing results in the communication sub-slot.

[0011] During the sensing phase, the UAV scans the target GT to obtain its corresponding azimuth and elevation angles. However, due to limitations in the scanning step size, there will be a certain quantization error between the estimated angle and the actual angle. Therefore, based on the geometric relationship between the detected direction vector of the UAV pointing towards the GT and the actual direction vector pointing towards the GT, this invention constructs a beam deflection angle model to derive the radial distance between the center of the FSO beam spot and the center of the GT receiving lens.

[0012] In the communication phase, the radial distance from the sensing phase is incorporated into the FSO channel model to describe the impact of pointing error on the communication link quality. The established FSO channel model includes system loss, atmospheric attenuation, atmospheric turbulence fading, and pointing error. Then, based on the FSO channel gain, the achievable rate model of GT is established, and on this basis, the average achievable rate model of GT over the entire mission cycle is constructed.

[0013] Based on the aforementioned UAV-assisted O-ISAC system model, sensing model, and communication model, this invention, under the constraints of maximum beam misalignment, sensing time, and UAV kinematics, maximizes the minimum average reachable rate among all ground planes by jointly optimizing UAV trajectory, ground terminal association scheduling, and sensing time allocation. The optimization problem is expressed as:

[0014]

[0015]

[0016]

[0017]

[0018]

[0019]

[0020]

[0021]

[0022] in, This is the set of correlation variables between UAV and GT. For the set of UAV locations, For the set of perceived time; For UAV and ground terminal The average achievable rate per time slot; This means that at most one ground terminal can be served in each time slot; Indicates in time slot Does the UAV serve the ground terminal? ; For FSO beam at ground terminal From the center of the light spot to the ground terminal Radial distance from the center of the lens; For FSO beam at ground terminal The beam width at that location; Sensing sub-slot duration, For communication sub-slot duration, The duration of the time slot; This is the maximum flight speed of the UAV; and These are the starting and ending positions of the UAV, respectively. and This represents the regional boundary.

[0023] Since the associated variables are integers and the reachability rate is nonlinear, the above problem is a mixed integer nonlinear programming problem. Therefore, an alternating iterative optimization algorithm based on the Block Coordinate Descent (BCD) method is proposed to solve it. The problem is decomposed into two subproblems, which are solved by combining Penalty Successive Convex Approximation (P-SCA), closed-form solutions, and SCA.

[0024] The advantages and beneficial effects of this invention are as follows:

[0025] This invention proposes a user-fairness-oriented UAV-assisted optical-sensing integrated resource allocation method. By jointly optimizing UAV trajectories, ground terminal association scheduling, and sensing time allocation, it maximizes the minimum average reachability rate among all ground terminals while satisfying constraints such as maximum beam misalignment, UAV flight speed, flight area, and sensing time. This effectively improves service fairness and system resource utilization efficiency in multi-user scenarios. Simulation results demonstrate that this invention can significantly improve the minimum average reachability rate and enhance multi-user communication fairness while ensuring beam alignment and UAV flight feasibility, demonstrating strong application value and development potential.

[0026] 1. Joint Modeling of Dynamic Sensing Error and FSO Pointing Error: To address the issue that existing research does not adequately consider beam deflection and receiver radial offset caused by sensing angle errors in UAV dynamic flight scenarios, this invention establishes the relationship between sensing angle error, beam deflection angle, and beam radial offset in claims 1 and 2 and step S1 of the specification. Furthermore, the radial offset is incorporated into the FSO pointing error loss and maximum beam misalignment constraint, thereby more accurately describing the impact of UAV dynamic position changes on optical communication link performance.

[0027] 2. Joint optimization for user fairness: Existing UAV-assisted O-ISAC scenario resource allocation methods often focus on the total system rate or energy consumption, which can easily lead to insufficient service for weak users. In claims 4-8 and steps S2-S3 of the specification, this invention jointly optimizes UAV trajectory, ground terminal association scheduling, and sensing time allocation to maximize the minimum average reachable rate among all ground terminals. For this mixed integer nonlinear programming problem, BCD combined with P-SCA, sensing time closed-form solution, and SCA is used for solution, thereby taking into account beam alignment, flight feasibility, and multi-user fairness.

[0028] This invention combines the improvements mentioned above. First, based on the UAV's dynamic position and sensing angle error, beam deflection and receiver radial offset are calculated and used for FSO channel gain and beam misalignment constraint modeling. Then, considering the interplay between UAV trajectory, associated scheduling, and sensing and communication times, a joint optimization model is constructed. By alternately solving for associated scheduling, sensing time, and UAV trajectory, the minimum average reachability of the system is improved while ensuring beam alignment and UAV flight constraints. Attached Figure Description

[0029] Figure 1 This is a schematic diagram of the system model structure of a preferred embodiment of the present invention;

[0030] Figure 2 This is a schematic diagram of the time slot structure of the present invention;

[0031] Figure 3 This is a flowchart of the present invention;

[0032] Figure 4 Convergence graphs of the algorithm under different ground truth (GT) numbers and task cycles;

[0033] Figure 5 UAV trajectory diagrams under different mission cycles;

[0034] Figure 6 A graph showing the relationship between task duration and minimum average achievable rate;

[0035] Figure 7 This is a graph showing the relationship between the number of ground truths (GTs) and the minimum average reachable rate. Detailed Implementation

[0037] The technical solutions of the embodiments of the present invention will be clearly and thoroughly described below with reference to the accompanying drawings. The described embodiments are merely some embodiments of the present invention.

[0038] The technical solution of the present invention to solve the above-mentioned technical problems is:

[0039] To address the aforementioned problems, this invention proposes a user-fairness-oriented UAV-assisted integrated optical and sensing resource allocation method. Under the premise of satisfying maximum beam misalignment constraints, UAV flight speed constraints, flight area constraints, and sensing time constraints, this method maximizes the minimum average reachable rate among all ground terminals by jointly optimizing UAV flight trajectories, ground terminal association scheduling, and sensing time allocation. To achieve the above objective, this invention employs the following... Figure 2 The technical solution shown is implemented as follows:

[0040] S1: Construct a UAV-assisted O-ISAC system model, and further construct a sensing model and a communication model based on the UAV-assisted ISAC system transmission model.

[0041] Includes the following steps:

[0042] S11: Constructing a UAV-assisted O-ISAC system model

[0043] System model structure as follows Figure 1 As shown. Consider a UAV-assisted O-ISAC network in a post-disaster emergency communication scenario, which includes a UAV and GT, composed of set The text indicates that the GT (Ground Zero) comprises sparsely distributed rescue nodes such as the core rescue terminal and temporary medical points, with the UAV (Unmanned Aerial Vehicle) acting as an airborne base station transmitting large volumes of data to the GT. Before the mission, the UAV has pre-obtained the coordinates of all GTs via GPS. However, FSO (Free-Side Array) communication has stringent pointing requirements; to ensure communication quality, the UAV must acquire precise pointing information in real time to aid in the alignment of the communication beam. Assuming the UAV operates at a constant altitude... During the task cycle The ISAC task is executed within the entire task cycle. Decomposed into equal time intervals of The set of time slots is represented as Each time slot is further subdivided into sensing sub-time slots. and communication sub-slots , Time slot structure such as Figure 2 As shown. In the sensing sub-time slot, the UAV acquires the azimuth and elevation angles of the GT being served. Subsequently, in the communication sub-time slot, the UAV uses the sensing information obtained in the sensing phase to perform FSO beam alignment, thereby improving the downlink transmission rate of the FSO link. If small enough, the UAV's trajectory can be approximated as a relatively short straight line within each time slot. (In the time slot...) The location coordinates of the UAV can be represented as: ,in These are the horizontal coordinates of the UAV.

[0044] S12: Establish a perception model

[0045] In a sensing system, assume GT The position coordinates are Because the system's ranging error is very small, and much smaller than the distance between the UAV and GT, the measured distance between the UAV and GT is... The distance is expressed as:

[0046] (1)

[0047] Let the pitch angle between the UAV and the GT be . azimuth angle is UAV points to GT unit direction vector It can be represented as:

[0048] (2)

[0049] Due to horizontal scan step size and vertical scan step size Due to limitations, the pitch and azimuth angles estimated by the UAV contain quantization errors. Therefore, the estimated GT... The elevation and azimuth angles of the lens center can be expressed as: , ; , .in For pitch angle resolution, This represents the resolution of the azimuth angle. Assume the scanning range in the vertical direction is... The scanning frequency is The time required for scanning is Then the resolution of the pitch angle is ,when and When fixed, along with The direction vector of the actual emitted FSO beam decreases as the number of beams increases. for

[0050] (3)

[0051] Define UAV to GT The center of the FSO spot and the UAV to GT The angle between the centers of the lenses is the beam deflection angle. The distance from the center of the light spot to GT The radial distance from the center of the lens can be expressed as:

[0052] (4)

[0053] According to Theorem 1, the square of the radial distance is:

[0054] (5)

[0055] Theorem 1: The distance from the center of the light spot to GT The radial distance from the center of the lens is:

[0056] (6)

[0057] Proof: According to the dot product formula, the cosine of the beam deflection angle is equal to... and dot product:

[0058] (7)

[0059] Substitute the coordinates and expand. It can be represented as:

[0060] (8)

[0061] According to the trigonometric difference formula It can be simplified to

[0062] (9)

[0063] Due to pitch angle resolution and azimuth resolution It's very small, so we can use a small angle approximation. , ,therefore

[0064] (10)

[0065] (11)

[0066] The two approximations and Substitute Expanding and ignoring minterms, then simplifying, we get:

[0067] (12)

[0068] Because of the beam deflection angle Since it is extremely small, its own cosine value can also be approximated, and we can obtain:

[0069] (13)

[0070] From the center of the light spot to GT The radial distance from the center of the lens can be expressed as:

[0071] (14)

[0072] Due to pitch angle The resolution of the pitch angle is Substituting into the above equation, we get:

[0073] (15)

[0074] Because this sensing method employs a multi-line scanning mode, its structure is similar to LiDAR, with the transmitter emitting multiple laser beams via a spatial light modulator. Interference between the sensing beams can lead to reduced SNR of the surrounding signal and ghost target issues. Therefore, the adjacent divergence angles of the multiple beams are crucial. Single beam divergence angle twice as much as To avoid beam overlap and blind spots, we will Set to:

[0075] (16)

[0076] Due to the narrow beam characteristics of FSO, a UAV can only perform an ISAC task for one GT within a time slot. A binary variable is introduced. If UAV is in time slot For GT Executing the ISAC task, then ,otherwise .

[0077] The performance of FSO communication is extremely sensitive to beam alignment; beam misalignment (i.e., the beam spot deviating from the receiving lens) will lead to a sharp decline in communication quality. To control this pointing error caused by sensing inaccuracies within an acceptable range, we introduce a maximum beam misalignment constraint, which requires that in any service time slot... In the middle, radial distance All must be less than the beam width of the FSO beam at GT k. ,Right now:

[0078] (17)

[0080] S13: Establish a communication model

[0081] S131: Establish FSO channel model

[0082] The channel model based on FSO mainly includes transceiver system losses, atmospheric attenuation, attenuation caused by turbulence, and pointing error, UAV and GT. The FSO channel model for inter-channel communication links can be represented as:

[0083] (18)

[0084] in For system efficiency, , and These are atmospheric attenuation, atmospheric turbulence attenuation, and pointing error loss, respectively.

[0085] Atmospheric attenuation loss includes absorption loss and scattering loss. Absorption loss occurs when photons in the FSO beam collide with gas molecules, but it is negligible compared to the scattering loss that occurs when the FSO beam collides with atmospheric particles. Scattering loss depends on weather factors such as snow, rain, and fog. The scattering loss of the FSO signal can be expressed using the Beer-Lambert law as follows:

[0086] (19)

[0087] in It is the extinction coefficient, expressed as:

[0088] (20)

[0089] in It's visibility. This is a reference value for the visible light range (550nm). It is the FSO wavelength. The scattering size distribution follows the Kruse model:

[0090] (twenty one)

[0091] The atmospheric turbulence decay between the UAV and GT in this paper is mainly caused by weak turbulence at low altitudes. Therefore, the atmospheric turbulence can be described by a log-normal model, and the probability density function (PDF) is given by the following equation:

[0092] (twenty two)

[0093] in and Let be the mean and variance of the logarithmic radiation intensity. However, The variance is very small. Based on the conventional UAV network parameters [5], it can be assumed that The average value is 1. Therefore, this paper ignores the effect of turbulence.

[0094] The pointing error loss based on FSO can be expressed as [6]:

[0095] (twenty three)

[0096] in It is the distance from the center of the FSO spot to the GT. Radial distance of the lens center For GT The received power at that location, It is the FSO beam in GT The beam radius at the receiving aperture. Assuming the receiving aperture radius is much smaller than the beam radius at the receiving end, the beamwidth... It can be approximated as being proportional to the distance. Therefore and They can be represented as , ,in It is the beam divergence angle. The radius of the receiving aperture.

[0097] S132: Establish an average reachability model

[0098] Considering variance UAV with zero mean Gaussian noise and GT The reachable rate between them can be expressed as:

[0099] (twenty four)

[0100] in For the bandwidth of the FSO link, , For the UAV transmit power, c is a parameter related to the average and peak power allowed by the UAV. Therefore, GT The average achievable rate per time slot is expressed as:

[0101] (25)

[0102] S2: Establish a system optimization model to maximize the minimum average reachable rate.

[0103] Constructing the aforementioned communication and sensing models, this invention proposes, under constraints of maximum beam misalignment, UAV flight speed, flight area, and sensing time, to maximize the minimum average reachability among all ground terminals by jointly optimizing UAV trajectories, associated scheduling, and sensing time allocation. The problem is expressed as:

[0104] (26)

[0105]

[0106]

[0107]

[0108]

[0109]

[0110]

[0111]

[0112] in, This is the set of correlation variables between UAV and GT. For the set of UAV locations, For the set of perceived time; For UAV and ground terminal The average achievable rate per time slot; This means that at most one ground terminal can be served in each time slot; Indicates in time slot Does the UAV serve the ground terminal? ; For FSO beam at ground terminal From the center of the light spot to the ground terminal Radial distance from the center of the lens; For FSO beam at ground terminal The beam width at that location; Sensing sub-slot duration, For communication sub-slot duration, The duration of the time slot; Maximum flight speed of the UAV; UAV starting position, UAV ending position, and area boundary;

[0113] S3: Design an alternating optimization algorithm based on BCD to solve the optimization problem.

[0114] Due to the problem The non-convexity of the objective function makes direct solution difficult. To overcome this problem, auxiliary variables are introduced. This makes the problem This can be equivalently restated as:

[0115] (27)

[0116]

[0117]

[0118] Since the correlation variables between UAV and GT are integer variables, and the achievable rate is non-linear, the problem... This is a mixed-integer nonlinear programming (MINLP) problem. Therefore, an alternating iterative optimization algorithm based on the BCD method is proposed to solve it. We can decompose (P0-1) into two sub-problems. Sub-problem one is to optimize only the UAV scheduling given the sensing time and UAV trajectory; sub-problem two is to optimize the UAV trajectory and sensing time given the UAV scheduling.

[0119] S31: Solving the UAV and GT correlation optimization

[0120] Given the sensing time and UAV trajectory, the UAV task scheduling optimization subproblem can be written as:

[0121] (28)

[0122]

[0123] Due to the binary nature of UAV scheduling variables, the problem This is a non-convex problem. To solve this optimization problem, the non-convex binary constraint C2 is equivalently transformed into the following constraint:

[0124] (29)

[0125] (30)

[0126] The equivalence proof is as follows: For any , It always holds true; in addition, satisfying equation (19) means that or Therefore, in conjunction with equation (30), if and only if or When, equation (30) holds, that is .question This can be equivalently restated as:

[0127] (31)

[0128]

[0129]

[0130]

[0131] Given the sensing time and UAV trajectory, it can be seen that, except for constraint C6, the objective function and constraints are convex. Therefore, the problem... The problem remains non-convex, making it difficult to solve. Generally, we can use a first-order Taylor series expansion to transform the non-convex constraints into convex ones, and then use the SCA method to iteratively solve the transformed optimization problem. However, directly applying SCA to solve the problem... Equations (29) and (30) lead to an infeasible solution. This problem is overcome by transforming constraint C10 into an objective function using a penalty method. It can be rephrased as:

[0132] (32)

[0133]

[0134] in As a penalty factor. Let In the initial stage, a smaller penalty factor can be set. To expand the feasible region; then gradually increase it. The value forces It tends to 0. When At that time, the problem and Equivalent. At this point, we can use the SCA method to handle the non-convex constraint C6.

[0135] As you can see, Since it is non-convex, its global upper bound can be obtained by performing a first-order Taylor expansion:

[0136] (33)

[0137] in For iteration variables, In the first The local point obtained in the next iteration. Problem It can be rephrased as:

[0138] (34)

[0139]

[0140] According to (34), It is convex. Therefore, this problem can be effectively solved by using existing convex optimization toolkits, such as CVX. By using SCA, the minimum average achievable rate is continuously updated until it converges to a stable value.

[0141] S32: UAV trajectory and perception time optimization

[0142] For a given UAV schedule, the joint optimization of the UAV trajectory and sensing time subproblem can be written as:

[0143] (35)

[0144]

[0145] because In It is a highly nonconvex function. To transform it into a more tractable function, we use a logarithmic approximation, which provides a tight lower bound on the original function, and we can utilize the 1st... The optimization result of the second iteration is used to approximate the third iteration. The function for the next iteration:

[0146] (36)

[0147] When the approximate constants are chosen in the following manner, the inequality applies. The condition is tight (i.e., the inequality takes the equality sign).

[0148] , (37)

[0149] Therefore, we obtain the above method. The lower bound is denoted as

[0150] (38)

[0151] in .

[0152] question It can be written as:

[0153] (39)

[0154]

[0155]

[0156] S321: Optimization of Perception Time

[0157] Given a set of drone locations In this case, the problem It can be written as:

[0158] (40)

[0159]

[0160] To facilitate problem solving, in (38) Constructed in the form of perceiving time:

[0161] (41)

[0162] in ,

[0163] question It can be rewritten as:

[0164] (42)

[0165]

[0166]

[0167] The LHS of constraint C12 is a concave function and can be reduced to

[0168] (43)

[0169] in , , .

[0170] Due to the problem It is a convex optimization problem. By solving the optimization problem, we can obtain the solution. The closed-form solution is shown in Theorem 2.

[0171] Theorem 2: The optimal sensing time is:

[0172] (44)

[0173] Proof: Let , The first derivative is:

[0174] (45)

[0175] The second derivative is:

[0176] (46)

[0177] when , hour, It is always less than 0, therefore It is monotonically decreasing. When hour, ;because It is monotonically decreasing. ,if ,at this time It is always less than 0 within its domain, therefore The maximum value is ,at this time unreasonable.

[0178] To make , must meet ,at this time There exists a unique zero point within the domain. It has a maximum value. According to Shengjin's formula, its unique real root in the range can be obtained:

[0179] (47)

[0180] in .

[0181] therefore The maximum point is Optimal perception time .

[0182] S321: UAV trajectory optimization

[0183] To facilitate problem solving, Constructed in the form of UAV trajectories:

[0184] (48)

[0185] in , , , , It can be seen that... , .also This is explained in Theorem 3.

[0186] Theorem 3: .

[0187] Proof: Because ,like ,but ,Right now .make Substitute ,have to .because , ,so ,Right now .therefore .

[0188] Now, substituting (38) and (48) into constraint C8, constraint C8 is still non-convex. In order to transform it into a convex constraint, slack variables are introduced. , , The problem is transformed into

[0189] (49)

[0190]

[0191]

[0192]

[0193]

[0194]

[0195] Due to the non-convex constraints C13, C14, and C15, the problem This remains a nonconvex problem. To address these difficulties, we apply a similar SCA technique to obtain the lower bound of the RHS first-order Taylor expansion of C14.

[0196] (50)

[0197] Replacing the lower bound of the constraint with the RHS of C14, C14 can be expressed in the following form.

[0198] (51)

[0199] For the non-convex constraint C15, taking the exponents on both sides of the inequality, C15 becomes...

[0200] (52)

[0201] Since both LHS and RHS are convex functions, this constraint is still considered non-convex. Therefore, a first-order Taylor expansion of the RHS is performed.

[0202] (53)

[0203] Constraint C15 is transformed into

[0204] (54)

[0205] For the non-convex constraint C16, a similar method is used, and C16 is transformed into:

[0206] (55)

[0207] question It can be restated in the following approximate form:

[0208] (56)

[0209]

[0210] At this point, the problem This is a convex optimization problem, in which the CVX solver can be used to solve it efficiently. By solving the sensing time and UAV trajectory problems, a suboptimal solution was eventually obtained.

[0211] Based on the above analysis and discussion, a BCD algorithm is proposed to iteratively solve the subproblems for solving the main problem P0 until the maximum and minimum fairness gap between two iterations is less than a threshold or the number of iterations is greater than the maximum number of iterations, as shown in Algorithm 1.

[0212] Table 1. Resource Allocation Algorithm Based on BCD

[0213]

[0214] Solve and Both use convex solvers based on the neighborhood point method, therefore their computational complexities are respectively , ,in It is a CVX solution. The number of iterations required for convergence It is a CVX solution. The number of iterations required for convergence. This refers to the convergence accuracy. The number of iterations required for Algorithm 1 to converge is... At that time, solve the main problem The time complexity of Algorithm 3 is .

[0215] To demonstrate the effectiveness of the proposed scheme, a comprehensive simulation evaluation was conducted on the effectiveness of the proposed BCD-based alternating iterative algorithm in a UAV-assisted FSO-ISAC network. The algorithm's performance was compared with three benchmark schemes to verify its effectiveness.

[0216] 1) Comparison with Scheme 1 (fixed circular trajectory): The association and perception time of the UAV are optimized, but the UAV flies from the starting point to the end point along a preset circular trajectory at a constant speed.

[0217] 2) Comparison with Scheme 2 (Greedy Association): The flight trajectory and sensing time of the UAV are optimized, but the UAV is preferentially associated with the ground terminal with the best channel conditions.

[0218] 3) Comparison with Scheme 3 (fixed time ratio): The flight trajectory and association of UAVs are optimized, but the ratio of perception time to communication time remains unchanged.

[0219] In addition, the main simulation parameters are listed in Table 2 and remain unchanged unless otherwise stated.

[0220] Table 2 Simulation Parameter Settings

[0221]

[0222] By comparing the simulation results with the above schemes, this example provides a specific simulation and the conclusions drawn.

[0223] Figure 4 The study presents the trend of the minimum average reachability rate (MINAR) with the number of iterations under different task durations T and different ground truth (GT) numbers K. It can be observed that the MINAR gradually increases with the number of iterations, reaching a stable convergence state after approximately four iterations, confirming the proposed algorithm's fast convergence performance. Furthermore, shorter task durations lead to a decrease in the MINAR. Similarly, a larger number of GTs also results in a decrease in the MINAR, because shorter task durations and larger GT numbers intensify resource competition, reducing the time the UAV spends serving each GT, thus degrading overall performance.

[0224] from Figure 5The image shows the UAV trajectories under different mission periods T. It can be observed that as the mission period T increases, the UAV's trajectory gradually approaches the ground plane (GT). Especially when T is 50 seconds, its trajectory can generally closely approximate all GTs within the region. This is because a larger T provides a greater flight time budget, allowing the UAV to design a longer route to actively approach the GT, thereby improving channel conditions and ultimately increasing the system's minimum average rate.

[0225] Figure 6 The variation of the minimum average achievable rate with the task period T is shown. It can be seen that the minimum average achievable rate of all schemes increases with increasing T. This is because, with the number of time slots remaining constant, the duration of a single time slot increases with the increase of the task period, providing more available time for data transmission and thus improving the rate. The proposed scheme consistently performs optimally, and the gap between it and other schemes gradually widens with increasing T. Combined with... Figure 5 Analysis shows that: compared to schemes without trajectory optimization, the proposed scheme utilizes increased time resources to plan a flight path closer to the ground plane (GT), improving channel quality between the UAV and the GT; compared to fixed sensing time schemes, the proposed scheme can adaptively adjust sensing time according to the environment, increasing the minimum average reachable rate while meeting sensing performance requirements; compared to greedy correlation schemes, the proposed scheme, through joint optimization, coordinates trajectory planning and task allocation, enabling the UAV to actively approach the bottleneck GT to compensate for path loss, avoid the bottleneck effect, and achieve more balanced and efficient communication coverage.

[0226] Figure 7 The variation of the minimum average reachability rate with the number of ground truth (GT) K is shown. It can be observed that as the number of GTs increases, the minimum average reachability rate of all schemes decreases, and this decrease gradually flattens. This is because as the number of GTs increases, resource competition among GTs becomes more intense, and limited time resources need to be shared among more GTs, thus leading to a decrease in the minimum average reachability rate. The minimum average reachability rate of the proposed scheme is consistently higher than other benchmark schemes. When K = 12, it is 11%, 21%, and 89% higher than the circular trajectory, fixed sensing time, and greedy association schemes, respectively. This is because the proposed scheme jointly optimizes trajectory planning, task scheduling, and sensing time, solving the problem of conflict between sensing and communication resources and achieving efficient resource utilization. Compared with the circular trajectory scheme, the performance advantage of the proposed scheme becomes more significant with the increase of the number of GTs. This is because the GT distribution is relatively sparse, and the proposed scheme can better accommodate the communication needs of all GTs through trajectory optimization. Compared with the other three schemes, the scheme lacking association optimization consistently has the lowest minimum average reachability rate, indicating that the association strategy is of great significance for improving system performance. This invention specifically includes:

[0227] Existing UAV-assisted O-ISAC systems suffer from insufficient modeling of pointing errors under dynamic UAV flight conditions, making it difficult to guarantee communication fairness in multi-user scenarios due to resource contention. This invention proposes a UAV-assisted integrated optical-sensing resource allocation method oriented towards user fairness. This method effectively combines the perception-assisted communication mechanism with UAV trajectory planning, associated scheduling, and perception time allocation. In the perception phase, the UAV scans and senses to acquire the azimuth and pitch angle information of the served ground terminals. Based on the perception angle error, beam deflection angle, and receiver radial offset geometry, a pointing error model under dynamic trajectory is established. In the communication phase, pointing error loss is introduced into the FSO channel model to establish a ground terminal reachable rate model, characterizing the coupling relationship between perception time, pointing error, and communication performance. By satisfying constraints such as task scheduling, perception time, UAV flight speed, flight area, and maximum beam misalignment, the method jointly optimizes the UAV trajectory, UAV-ground terminal associated scheduling variables, and perception time allocation for each service time slot, maximizing the minimum average reachable rate among all ground terminals.

[0228] This invention can be applied to the field of unmanned aerial vehicle (UAV) communication, and is suitable for scenarios such as disaster relief communication, temporary network coverage, and high-speed data transmission for sparsely populated ground terminals. This solution introduces a UAV-assisted FSO communication link, which leverages the flexibility and rapid aerial movement capabilities of UAVs to dynamically schedule aerial service nodes within the target area, thereby achieving wide-area communication service coverage. Simultaneously, it utilizes the high speed and strong electromagnetic interference resistance of FSO to provide high-bandwidth wireless access services to the target area. Furthermore, this invention utilizes angle information from the sensing process to effectively reduce the impact of beam pointing errors caused by the dynamic flight of UAVs on the FSO link, improving link reliability and system resource utilization efficiency. It can also be extended to applications such as emergency support for smart cities, personnel communication during field operations, communication support for temporary large-scale events, border patrol, data backhaul for environmental monitoring, and low-altitude integrated air-ground networks, showing promising engineering application prospects.

[0229] The systems, devices, modules, or units described in the above embodiments can be implemented by computer chips or entities, or by products with certain functions.

[0230] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0231] The above embodiments should be understood as illustrative only and not as limiting the scope of protection of the present invention. After reading the description of the present invention, those skilled in the art can make various alterations or modifications to the present invention, and these equivalent changes and modifications also fall within the scope defined by the claims of the present invention.

Claims

1. A user-fair, unmanned aerial vehicle (UAV)-assisted integrated optical and sensor resource allocation method, characterized in that, Includes the following steps: S1. Construct an integrated optical-sensing O-ISAC system model assisted by a UAV, and build a sensing model and a free-space optical (FSO) communication model based on the system model. The system model is an emergency communication network where a single UAV serves K ground GT terminals. The total mission cycle is divided into N equal-length time slots, and each time slot is further divided into a sensing sub-time slot and a communication sub-time slot. In the sensing sub-time slot, the UAV obtains the azimuth and pitch angles of the target terminal by scanning. In the communication sub-time slot, the UAV adjusts the FSO beam direction based on the sensing results to complete downlink transmission. The UAV maintains a fixed flight altitude, and its flight trajectory within a single time slot is approximately a straight line. The perception model derives the beam deflection angle based on the scanning quantization error, solves the radial distance between the beam spot center and the receiving lens by combining vector geometry, and sets a maximum radial distance constraint to control beam misalignment; a multi-line LiDAR-type scanning scheme is adopted, and the divergence angle of adjacent perception beams is twice that of a single beam to avoid beam overlap and detection blind spots. The FSO communication model integrates four types of loss terms: system loss, atmospheric attenuation, weak atmospheric turbulence, and pointing error, and establishes a single-slot average achievable rate expression based on channel gain. S2. Construct a fairness optimization mathematical model: The optimization objective is to maximize the minimum average reachability of all ground terminals. The model jointly optimizes the UAV time slot position trajectory, the binary correlation variables between the UAV and the ground terminal, and the duration of each time slot's sensing sub-time slot. The constraints include that each time slot serves only one terminal, the radial distance is less than the beam width, the sensing duration and communication duration are equal to the total time slot length, the UAV's start and end positions and maximum flight speed, and regional boundary constraints. The optimization model is a mixed-integer nonlinear programming MINLP problem. S3. Design an alternating iterative algorithm based on block coordinate descent (BCD) to decompose and solve the optimization model. The original problem is decomposed into an associated scheduling subproblem and a joint optimization subproblem of trajectory and perception time, which are iterated alternately until convergence. S4. Iterative convergence determination: The two types of sub-problems are solved alternately. When the difference between the objective functions of the two iterations is less than the preset convergence accuracy or the number of iterations reaches the maximum threshold, the iteration stops and the optimal UAV trajectory, terminal association strategy, and time slot perception allocation duration are output.

2. The user-fairness-oriented UAV-assisted integrated optical and sensor resource allocation method according to claim 1, characterized in that, The radial distance derivation process in step S1 of the sensing model includes small-angle approximation: the trigonometric functions are simplified by utilizing the extremely small pitch and azimuth resolution characteristics, and the analytical expression of the radial distance of the light spot that dynamically changes with the position of the UAV is derived by combining the dot product formula of the beam vector; the pitch resolution of the sensing scan is determined by the vertical scanning range, scanning frequency, and sensing duration.

3. The user-fairness-oriented UAV-assisted integrated optical and sensor resource allocation method according to claim 1, characterized in that, In step S1 of the FSO communication model, atmospheric attenuation is calculated using the Beer-Lambert law combined with the Kruse scattering size model; weak low-altitude atmospheric turbulence is calculated using a log-normal model, ignoring turbulence variance fluctuations; pointing loss is determined by the radial distance of the light spot, the radius of the receiving light spot, and the radius of the receiving aperture.

4. The user-fairness-oriented UAV-assisted integrated optical and sensor resource allocation method according to claim 1, characterized in that, The specific steps of step S2, constructing a fairness optimization mathematical model, involve: under the constraints of maximum beam misalignment, UAV flight speed, flight area, and perception time, maximizing the minimum average reachability among all ground terminals by jointly optimizing UAV trajectories, associated scheduling, and perception time allocation. The problem can be expressed as: ； in, This is the set of correlation variables between UAV and GT. For the set of UAV locations, For the set of perceived time; For UAV and ground terminal The average achievable rate per time slot; This means that at most one ground terminal can be served in each time slot; Indicates in time slot Does the UAV serve ground terminals? ; For FSO beam at ground terminal From the center of the light spot to the ground terminal Radial distance from the center of the lens; For FSO beam at ground terminal The beam width at that location; In order to sense the duration of sub-slots, For communication sub-slot duration, The duration of the time slot; This is the maximum flight speed of the UAV; Let x be the x-coordinate of the UAV in the nth time slot. The maximum boundary length of the UAV's flyable region in the x-axis direction. Let the ordinate of the UAV in the nth time slot be... This represents the maximum boundary length of the UAV's flyable region along the y-axis.

5. The user-fairness-oriented UAV-assisted integrated optical and sensor resource allocation method according to claim 1, characterized in that, The S3 design uses an alternating iterative algorithm based on block coordinate descent (BCD) to decompose and solve the optimization model. The original problem is broken down into an associated scheduling subproblem and a joint trajectory and perception time optimization subproblem, which are iterated alternately until convergence. Specifically: S31. Correlation scheduling subproblem: Fix the UAV trajectory and perception duration, introduce penalized continuous convex approximation P-SCA combined with continuous convex approximation SCA iteration, relax the binary correlation constraint and approximate the integer constraint through the penalty term, perform first-order Taylor convexization approximation on the non-convex rate constraint, and use a convex solver to iteratively solve to obtain the optimal terminal correlation scheme. S32, Joint Optimization Subproblem of Trajectory and Sensing Time: Fix the terminal-related variables and use logarithmic tight lower bound approximation decomposition for the non-convex rate function; first, solve the closed-form optimal solution for sensing time given the UAV position; then, introduce relaxation variables for trajectory-related non-convex constraints, use SCA first-order expansion to convexize the solution, and output the UAV position for each time slot through a convex solver.

6. The user-fairness-oriented UAV-assisted integrated optical and sensor resource allocation method according to claim 5, characterized in that, The P-SCA processing method in step S31 is as follows: the penalty factor is gradually increased during the iteration process, forcing the relaxed correlation variable to approach the integer value of 0 / 1; the local expansion point is continuously updated and iteratively optimized for the non-convex constraint of reachable rate until the scheduling scheme converges.

7. The user-fairness-oriented UAV-assisted integrated optical and sensor resource allocation method according to claim 5, characterized in that, The derivation method of the closed-form solution for the sensing time in step S32 is as follows: the lower bound of the rate is transformed into a single-variable concave function of the sensing duration, the monotonicity of the function is determined by differentiation, and the unique optimal analytical solution for the sensing time is solved using the Shengjin formula.

8. The user-fairness-oriented UAV-assisted integrated optical and sensor resource allocation method according to claim 1, characterized in that, In step S32, during the UAV trajectory optimization process, relaxation variables are introduced for exponential non-convex constraints and fractional rate constraints, and first-order Taylor lower / upper bound convex approximations are made. After all constraints are converted into standard convex constraints, the CVX-type convex toolbox is used to solve them.

Images