An IRS quantity-aware networked unmanned aerial vehicle communication multi-dimensional resource optimization method

Abstract

This invention provides a multi-dimensional resource optimization method for networked unmanned aerial vehicle (UAV) communication based on IRS quantity awareness, specifically including the following steps: initializing base station information, CUAV information, and IRS information; establishing a system channel model, energy consumption model, and IRS quantity-aware CUAV downlink communication model based on the base station information, CUAV information, and IRS information; establishing a joint optimization problem model for maximizing system EE and minimizing the number of associated IRSs based on the system channel model, energy consumption model, and IRS quantity-aware CUAV downlink communication model; jointly optimizing the IRS association scheduling matrix, base station transmit power matrix, and IRS phase shift matrix to achieve the optimization objective corresponding to solving the joint optimization problem model for maximizing system EE and minimizing the number of associated IRSs; this invention can maximize system EE and minimize the number of associated IRSs while satisfying constraints such as communication connectivity, mobility, and unit modulus.

Description

Technical Field

[0001] This invention relates to the field of connected drone technology, and in particular to a multi-dimensional resource optimization method for connected drone communication based on IRS quantity awareness. Background Technology

[0002] Unmanned aerial vehicles (UAVs), with their superior performance in terms of flexibility, portability, and rapid response, have made UAV-assisted communication a research hotspot and cutting-edge direction in the field of wireless communication. However, existing UAVs mainly rely on simple point-to-point communication on unlicensed frequency bands, which suffer from limited spectrum resources, susceptibility to electromagnetic interference, and a lack of unified regulation. CUAV refers to integrating UAVs in various applications as new types of airborne user terminals into existing and future cellular network architectures, thereby achieving low-altitude intelligent connectivity through cellular communication technology.

[0003] However, traditional terrestrial base stations (BSs) often employ a downtilt deployment strategy to maximize communication coverage and quality for ground users. This significantly reduces signal coverage in low-altitude areas, leading to coverage blind spots and impacting CUAV communication connectivity. Secondly, while the air-to-ground link between CUAV and BSs is primarily line-of-sight (LoS) transmission, the air-to-ground channel suffers from fading. Considering the use of Ricean channel characterization, communication stability is significantly affected. Furthermore, the LoS-based air-to-ground channel also exposes CUAV to signal interference from numerous unrelated BSs, resulting in significant co-channel interference and a substantial deterioration in communication quality.

[0004] Intelligent reflectors (IRS) can dynamically adjust the amplitude and phase of incident signals, thereby enabling artificial control of the radio propagation environment, and have attracted widespread attention in recent years. By optimizing the phase response of each reflective element, IRS can reorient and reconstruct incident electromagnetic waves to the target direction, thereby enhancing the desired signal or suppressing interference signals. Therefore, it is expected to become a key technology for solving the core challenges of CUAV air-to-ground communication. However, current research on IRS-enabled CUAV air-to-ground communication systems is still in its early stages.

[0005] Although some studies have explored resource allocation methods for IRS-enabled UAV air-to-ground communication systems to enhance UAV communication transmission rates or reduce system energy consumption, many challenges remain to be addressed in resource allocation for IRS-enabled CAUV air-to-ground communication. These challenges mainly include: 1) Unlike UAV air-to-ground communication, CUAV, as an airborne user, needs to maintain a stable communication connection with ground base stations during flight. However, under the existing base station deployment that is mainly for ground user coverage, it is difficult to guarantee the communication performance of CUAV communication system powered by a single IRS throughout the entire flight cycle. 2) While introducing multiple IRS to power the CUAV communication system can help improve the system's communication performance while ensuring the CUAV's communication connection during flight, the higher degree of airspace freedom brought by multiple IRS and the need for adaptive configuration of the number of IRSs for energy saving both increase the difficulty of solving the problem. Summary of the Invention

[0006] The purpose of this invention is to improve and innovate upon the shortcomings and problems existing in the background technology, and to provide a multi-dimensional resource optimization method for networked drone communication based on IRS quantity awareness.

[0007] This invention provides a multi-dimensional resource optimization method for networked unmanned aerial vehicle (UAV) communication based on IRS quantity awareness, specifically including the following steps: Step S1, Method begins; Step S2: Initialize base station information, CUAV information, and IRS information; wherein, the base station information includes the number of base stations deployed, altitude, coordinates, number of antennas equipped, and number of IRS deployed next to each base station; the CUAV information includes the CUAV flight trajectory; and the IRS information includes the IRS deployment altitude, coordinates, and number of reflective units equipped. Step S3: Based on base station information, CUAV information, and IRS information, establish a system channel model, energy consumption model, and IRS quantity-aware CUAV downlink communication model; based on the system channel model, energy consumption model, and IRS quantity-aware CUAV downlink communication model, establish a joint optimization problem model for maximizing system EE and minimizing associated IRS quantity. Step S4: Jointly optimize the IRS association scheduling matrix, base station transmit power matrix, and IRS phase shift matrix to achieve the optimization objective corresponding to the joint optimization problem model of maximizing system EE and minimizing the number of associated IRSs; Step S5: Output the optimal compromise value of the optimization objective corresponding to the joint optimization problem model of maximizing the system EE and minimizing the number of associated IRS.

[0008] A further embodiment is that step S4 specifically includes: Step S41: Given the BS transmit power matrix and the IRS phase shift matrix, iteratively optimize the IRS scheduling factor matrix; Step S42: Given the IRS phase shift matrix and the IRS scheduling factor matrix obtained in step S41, iteratively optimize the BS transmit power matrix; Step S43: Given the IRS scheduling factor matrix and BS transmit power matrix obtained from the above steps, iteratively optimize the IRS phase shift matrix; Step S44: Determine whether the changes in the optimization objectives corresponding to maximizing the system EE and minimizing the number of associated IRSs are less than a preset threshold. Or, does the current iteration count exceed the preset maximum iteration count? If any condition is met, proceed to step S5; otherwise, proceed to step S41.

[0009] A further approach is that step S3, which establishes the system channel model based on base station information, CUAV information, and IRS information, specifically includes: Acquiring time slots n , No. m Between the first BS and CUAV, the first Between the IRS and CUAV and the first m The BS and the first Channel gain between IRS , and , This represents the index of the k-th IRS next to the m-th BS; This indicates the number of antennas equipped in each BS. The number of reflective units equipped in the IRS is indicated by the following formula: ; ; ; , and They represent time slots respectively. n No. m The three-dimensional distance of the first BS-CUAV, the first The three-dimensional distance of the first IRS-CUAV and the first m The BS-then Three-dimensional distance of each IRS; This represents the channel gain at a reference distance of 1m. , and These represent the path loss factors for the BS-CUAV link, IRS-CUAV link, and BS-IRS link, respectively. , and These represent the Rice factors for the BS-CUAV link, IRS-CUAV link, and BS-IRS link, respectively. , and They represent time slots respectively n , No. m Between the first BS and CUAV, the first Between the IRS and CUAV and the first m The BS and the first The deterministic LoS component of the channel gain between IRSs, and , and This corresponds to the NLoS component that follows a Rayleigh distribution; in, and The calculation formula is: ; ; in The working carrier wavelength, and These represent the BS antenna spacing and the IRS element spacing, respectively; and and They represent time slots respectively n , No. m The pitch departure angle and azimuth departure angle of the downlink from BS to CUAV satisfy the following relationship: ; and They represent time slots respectively n At that time, the first The pitch departure angle and azimuth departure angle of the downlink from the IRS to CUAV satisfy the following conditions: as well as Among them, the first m The BS and the first The horizontal coordinates of the IRS are respectively ;CUAV in the n The horizontal coordinate of the time slot is represented as ; Indicates the height of CUAV. Indicates the height of the IRS; at the same time, The calculation formula is: ; ; in, and They represent the first m BS to the first The pitch and azimuth angles of arrival of each IRS downlink, and , , ;in, Indicates the first m The BS and the first The array response vector at the IRS end between each IRS; Then it means the first m The BS and the first Between each IRS, the array response vector at the BS end, Indicates the height of BS.

[0010] A further approach is that step S3, which establishes the energy consumption model based on base station information, CUAV information, and IRS information, specifically includes: Among them, during the flight cycle The total energy consumption of the internal system is expressed as: ; Indicates time slot n No. m The transmit power of each BS; Indicates the time slot length; This indicates that CUAV is in flight cycle T The propulsion power consumption within the system is calculated using the following formula: ; In the formula, Indicates in time slot At that time, the flight speed of CUAV; and These represent the blade power and induced power under CUAV hovering conditions, respectively. , ; The rotor blade tip velocity, The average rotor induced velocity during hovering. and These represent air density and rotor disk area, respectively. and This indicates the fuselage drag ratio and rotor rigidity; Indicates the angular velocity of the blade; Indicates the rotor radius; Indicates the weight of the CUAV; Indicates the airfoil drag coefficient; This represents the incremental correction coefficient for induced power.

[0011] A further approach is that step S3, which establishes an IRS-aware CUAV downlink communication model based on base station information, CUAV information, and IRS information, specifically includes: Obtaining CUAV in the The achievable signal transmission rate for each time slot is calculated using the following formula: ; In the formula, Indicates the first Time slot, CUAV associated BS; B Indicates system communication bandwidth. Represents the square of the Euclidean norm; Indicates conjugate transpose; , and They represent time slots respectively n , No. Between the first BS and CUAV, the first Between the IRS and CUAV and the first The BS and the first Channel gain between IRS; Indicates CUAV and the first A binary scheduling factor associated with each IRS; Indicates the first The k-th IRS reflection matrix within the coverage area of ​​each base station; Indicates time slot n No. The transmit power of each BS; Indicates in time slot n The channel gain between the i-th BS and CUAV; Time slot n The transmit power of the i-th BS; M express The collection of ground-based BS; This represents the power of additive white Gaussian noise; For CUAV in the The expected achievable signal transmission rate of each time slot is obtained by calculating the following formula: ; in, ; ; In the formula, , , , , , , , ; This represents the pure Loss-of-Stake (LoS) channel gain from the i-th BS to the CUAV at time slot n. This represents the pure Loss channel gain from the (b[n], k)th IRS to the CUAV at time slot n. This represents the pure Loss channel gain from the b[n]th BS to the (b[n], k)th IRS at time slot n. This represents the expected value of the squared magnitude of the composite channel. This represents the pure NLoS channel gain from the i-th BS to the CUAV at time slot n. This represents the expected value of the squared modulus of the directly connected channel. , and These represent the equivalent reference gain of the Ricean channel LoS component for the BS-CUAV link, BS-IRS link, and IRS-CUAV link, respectively. , and They represent time slots respectively n No. The three-dimensional distance of the first BS-CUAV, the first The three-dimensional distance of the first IRS-CUAV and the first The BS-then The three-dimensional distance of each IRS.

[0012] A further approach is that step S3, which establishes a joint optimization problem model for maximizing system EE and minimizing associated IRS based on the system channel model, energy consumption model, and IRS number-aware CUAV downlink communication model, specifically includes: Construct the objective function to maximize the system's energy expenditure (EE) and the objective function to minimize the number of associated IRSs: ; Wherein, system EE can be represented as ; X represents the IRS associated scheduling matrix. Represents the BS transmit power matrix; Represents the IRS phase shift matrix; After normalizing the objective function, the multi-objective optimization problem is transformed into a single-objective optimization problem using a linear weighted sum method, resulting in a joint optimization problem model that maximizes the system's energy efficiency (EE) and minimizes the number of associated information sources (IRS). ; in, C 1 indicates the communication connection constraints when CUAV connects to BS. C 2- C 3 represents the scheduling factor constraint of the IRS. C 4 indicates that the BS's transmit power must not exceed its maximum transmit power budget. , C 5 represents the unit modulus constraint of the IRS; This represents the minimum transmission rate threshold for CUAV to maintain communication connectivity. It is an adjustable parameter. This represents the normalized system EE. This represents the number of non-associative IRS after normalization; Indicates the first The first IRS r Phase shift of each reflecting element; K express A collection of IRSs deployed next to each ground-based BS; Represents a set of time slots.

[0013] A further embodiment is that step S41 specifically includes: Given the BS transmit power matrix IRS phase shift matrix The joint optimization problem of maximizing system EE and minimizing the number of associated IRSs is simplified to:

[0014] ; The binary variables in problem (P1.1) Relaxation as a continuous variable , Indicates in time slot n When the scheduling factors of CUAV and the (b(n), k)th IRS communication are continuously associated, the model is reformulated as follows:

[0015] ; Introducing auxiliary variables and define ,Will In Components at a given local iteration point A first-order Taylor expansion is performed at the given location, and the SCA algorithm is used for approximation to obtain its global lower bound; where, for No. The value of the nth iteration; This represents the cascaded equivalent channel gain of the b[n]th BS to the (b[n],k)th IRS-CUAV at time slot n; but Represented as , This represents the lower bound of the desired transmission rate: ; in , ; ; This represents the expected power of the NLoS component of the directly connected link; This represents the expected power of the NLoS component in the cascaded link; This represents a linear approximation of the squared magnitude of the channel gain; The model then transforms into:

[0016] ; In the formula, This represents the total number of time slots. express The total number of IRS deployed next to each ground-based BS; and These respectively represent 0 IRS associated with each time slot and 1 IRS associated with each time slot. The system EE value of each IRS.

[0017] A further embodiment is that step S42 specifically includes: Given and The problem (P0) can be simplified to:

[0018] ; According to Tinkelbach's algorithm, Transform into ; in, , For optimal system EE parameters; ; ; In the formula, This represents the equivalent total transmission rate. This represents the logarithm of the average sum of the power of all received signals and noise during the flight time of CUAV; This represents the logarithm of the average sum of the power of the interference signals and noise received by CUAV during the flight time. Will At a given local iteration point A first-order Taylor expansion is performed at the given point, and the SCA algorithm is used for approximation to obtain its global upper bound. This indicates that at time slot n, the first... The BS transmit power value is obtained in the next iteration; the specific first-order Taylor expansion is: ; in, express A linear approximation of yields... An approximate lower bound: ; The model then transforms into:

[0019] .

[0020] A further embodiment is that step S43 specifically includes: For a given IRS scheduling factor matrix and BS transmit power matrix Problem (P0) simplifies to:

[0021] ; Introducing auxiliary variables and The SCA algorithm and SDR technology were used to process and obtain information about The standard convex optimization problem is then expressed by the model as follows:

[0022] ; in, This represents the sum of all elements on the main diagonal of the matrix; The autocorrelation matrix represents the equivalent complex gain matrix of the CUAV received signal. express The autocorrelation matrix of the augmented vector; , ;in, This represents the total number of IRS associated with CUAV under the b[n]th BS at time slot n; , , This represents the equivalent complex gain matrix of the CUAV received signal. express augmented vector; ; ; ; This represents the equivalent concatenated channel preprocessing matrix from BS to CUAV. This represents the matrix formed by combining the phase shift vectors of the associated IRS within the b[n]th BS under time slot n; This represents the set of IRS associated sequences communicating with CUAV within the range of the b[n]th associated BS at time slot n; Must meet and ; Indicates the nth time slot The Each element.

[0023] Compared with existing technologies, the beneficial effects of this invention are as follows: This invention addresses the CUAV air-to-ground downlink communication scenario and proposes a multi-dimensional resource optimization method for high-efficiency communication of networked UAVs with IRS quantity awareness, specifically for situations involving multiple IRS association selection. The aim is to maximize the system energy efficiency (EE) while minimizing the number of associated IRSs under constraints such as communication connectivity, mobility, and unit modulus. This invention provides a new design concept and method for improving the communication reliability and green flight of CUAVs in low-altitude patrol and monitoring. Specifically, for IRS-enabled CUAV air-to-ground communication systems, considering the complexity and cost of long-term system operation and maintenance, this invention avoids blindly associating all IRSs during CUAV flight. Instead, while satisfying connectivity constraints, it minimizes the total number of associated IRSs within the total mission duration and maximizes the system EE by jointly optimizing the IRS association scheduling matrix, BS transmit power matrix, and IRS phase shift matrix. Furthermore, to handle the established non-convex optimization problem, an efficient iterative optimization algorithm based on AO, SCA, SDR, and Tinkelbach algorithms is proposed. In summary, the multi-dimensional resource optimization method for enabling high-efficiency air-to-ground communication in CUAV using IRS can effectively achieve a trade-off between system EE and the number of associated IRSs, which is beneficial for improving the service performance of CUAV under limited resource constraints. Attached Figure Description

[0024] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0025] Figure 1 This is a schematic diagram of a multi-dimensional resource optimization method for networked drone communication based on IRS quantity awareness, provided in an embodiment of the present invention. Figure 2 This is the IRS-enabled CUAV air-to-ground cooperative communication system model provided in the embodiments of the present invention; Figure 3This is a comparison chart of the system EE convergence performance under different schemes provided in the embodiments of the present invention; Figure 4 This is an iterative convergence curve of the system EE and the number of associated IRS under different flight times and weighting factors provided in the embodiments of the present invention; Figure 5 This is a diagram illustrating the trade-off between the number of system EEs and associated IRSs provided in this embodiment of the invention. Figure 6 The number of system EE and associated IRS provided in the embodiments of the present invention varies with A diagram showing the relationship between changes; Figure 7 The embodiments of the present invention provide that in and The dynamic variation characteristics of SE under different schemes are shown in the figure. Detailed Implementation

[0026] To make the objectives, features, and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0027] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used herein in the description of the invention is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.

[0028] like Figure 1As shown, this invention provides a multi-dimensional resource optimization method for networked unmanned aerial vehicle (UAV) communication based on IRS quantity awareness, which consists of five steps: Step S1: Method start; Step S2: Initialize base station information, CUAV information, and IRS information; wherein, base station information includes the number of base stations deployed, altitude, coordinates, number of antennas equipped, and number of IRS deployed next to each base station; CUAV information includes CUAV flight trajectory; IRS information includes IRS deployment altitude, coordinates, and number of reflective units equipped; Step S3: Establish a system channel model, energy consumption model, and IRS quantity awareness CUAV downlink communication model based on base station information, CUAV information, and IRS information; establish a joint optimization problem model for maximizing system EE and minimizing associated IRS quantity based on the system channel model, energy consumption model, and IRS quantity awareness CUAV downlink communication model; Step S4: Jointly optimize the IRS association scheduling matrix, base station transmit power matrix, and IRS phase shift matrix to achieve the optimization objective corresponding to the joint optimization problem model for maximizing system EE and minimizing associated IRS quantity; Step S5: Output the optimal compromise value of the optimization objective corresponding to maximizing system EE and minimizing associated IRS quantity. The steps described above will be explained in detail below: Step S1: Method begins; Step S2: Initialize base station information, CUAV information, and IRS information; wherein, the base station information includes the number of base stations deployed, altitude, coordinates, number of antennas equipped, and number of IRS deployed next to each base station; the CUAV information includes the CUAV flight trajectory; and the IRS information includes the IRS deployment altitude, coordinates, and number of reflective units equipped. Specifically, such as Figure 2 As shown, in this invention, each time slot-associated BS transmits information to the CUAV via a direct connection channel. Simultaneously, multiple IRSs adjacent to the associated BS reflect signals to the CUAV via cascaded channels to enhance the desired received signal. However, due to the high probability of Loss of Sight (LoS) characteristics of the air-to-ground channel, the CUAV will also receive co-channel interference signals transmitted by non-associated BSs.

[0029] It should be noted that this invention is by M Ground BS, It consists of an IRS and a CUAV, with each ground BS deployed next to it. K Each IRS (In-Relationship System) allows the CUAV to follow a pre-planned optimal flight trajectory (which can be obtained by solving the EE maximization model of a networked UAV air-to-ground communication system with single IRS assistance) during its flight cycle. T From the starting point Fly to the destination Assume all ground-based base stations (BS) have the same height. And each BS is equipped with A uniform linear array of antennas, the height of CUAV is... Meanwhile, to improve the LoS link performance between the IRS and the BS and CUAV, all IRSs are deployed at a height of [missing information - likely a specific altitude]. The fixed building surface, and all of them are equipped with A uniform planar array of reflective elements, wherein and These represent the number of reflective elements in the IRS along the y-axis and z-axis directions, respectively; where the y-axis direction corresponds to the horizontal direction (e.g., transverse) of the reflective element, and the z-axis direction corresponds to the vertical direction (e.g., longitudinal) of the reflective element.

[0030] Using a three-dimensional Cartesian coordinate system, let Indicates the first m Next to BS k Index of IRS, , and define the first m The BS and the first The horizontal coordinates of the IRS are respectively , From the starting point Fly to the destination CUAV in the n The horizontal coordinate of the time slot is represented as , .

[0031] Step S3: Based on base station information, CUAV information, and IRS information, establish a system channel model, energy consumption model, and IRS quantity-aware CUAV downlink communication model; based on the system channel model, energy consumption model, and IRS quantity-aware CUAV downlink communication model, establish a joint optimization problem model for maximizing system EE and minimizing associated IRS quantity. 1) Constructing a system channel model: in time slots n , No. m Between the first BS and CUAV, the first Between the IRS and CUAV and the first m The BS and the first The channel gain between the IRSs is expressed as follows: , and Since the channels between CUAV and the ground BS and IRS are mainly LoS links with multiple scattering components, this invention assumes that all communication links follow the Ricean channel model. Therefore, the channel gains can be expressed as follows: (1) (2) (3) in Indicates time slot n No. m The three-dimensional distance of each BS-CUAV Indicates in time slot n Time The three-dimensional distance of each IRS-CUAV Indicates the first m The BS-then The three-dimensional distance of each IRS. Furthermore, This represents the channel gain at a reference distance of 1m. , and These represent the path loss factors for the BS-CUAV link, IRS-CUAV link, and BS-IRS link, respectively. , and These represent the Rice factors for the BS-CUAV link, IRS-CUAV link, and BS-IRS link, respectively. , and In the time slot n , No. m Between the first BS and CUAV, the first Between the IRS and CUAV and the first m The BS and the first The deterministic LoS component of the channel gain between IRSs, and , and This corresponds to the NLoS component that follows a Rayleigh distribution.

[0032] at the same time, and The time-varying characteristics follow a quasi-static block fading model. Since the duration of each time slot is much longer than the length of a fading block, each time slot is divided into... The first fading block. Specifically, the first... n The first time slot ， Within a decay block, and It remains constant, while different fading blocks are independent of each other. and They represent time slots respectively n Within the sl-th fading block, the th m Between the BS and CUAV and the first The NLoS component of the channel gain between the IRS and CUAV.

[0033] also, and It can be represented as: (4) (5) in The working carrier wavelength, and These represent the BS antenna spacing and the IRS element spacing, respectively. and They represent time slots respectively n , No. m The pitch departure angle and azimuth departure angle of the downlink from BS to CUAV satisfy the following relationship: Similarly, and They represent time slots respectively n At that time, the first The pitch departure angle and azimuth departure angle of the downlink from the IRS to CUAV satisfy the following conditions: as well as .

[0034] at the same time, It can be represented as: (6) (7) in, and They represent the first m BS to the first The pitch and azimuth angles of arrival of each IRS downlink, and , , .in, Indicates the relationship between the m-th BS and the th The array response vector at the IRS end between each IRS; This indicates that the m-th BS is related to the m-th BS. The array response vector at the BS end between each IRS.

[0035] 2) Constructing an energy consumption model: In this invention, the total system energy consumption includes the propulsion energy consumption of CUAV, the communication energy consumption of CUAV (including the energy consumption consumed by signal processing and other circuit operations), the transmission energy consumption of BS, and the energy consumption of IRS components.

[0036] Among them, the rotor CUAV during flight cycle T The internal propulsion power consumption is: (8) In the formula, Indicates in time slot At that time, the flight speed of CUAV; and These represent the blade power and induced power under CUAV hovering conditions, respectively. , . The rotor blade tip velocity, The average rotor induced velocity during hovering. and These represent air density and rotor disk area, respectively. and This indicates the fuselage drag ratio and rotor rigidity; Indicates the angular velocity of the blade; Indicates the rotor radius; Indicates the weight of the CUAV; Indicates the airfoil drag coefficient; This represents the incremental correction coefficient for induced power.

[0037] It should be noted that in practical applications, the power consumption of CUAV communication is usually less than 10 watts, and the power consumption of IRS is usually less than 0.1 watts, both of which are much smaller than the hundreds of watts of propulsion power consumption of CUAV. Therefore, this invention assumes that the influence of CUAV communication power consumption and IRS power consumption is negligible.

[0038] Therefore, during the flight cycle The total energy consumption of the system can be expressed as: (9) in, Indicates time slot n No. m The transmit power of each BS is determined; the transmit power value is then obtained through optimization. Indicates the time slot length.

[0039] 3) Construct an IRS-aware CUAV downlink communication model: Assume that each time slot CUAV can be dynamically selected and associated with an IRS, and define a binary scheduling factor for associating CUAV with the (m,k)th IRS. :when "Time" indicates a time slot. n CUAV and the The first BS within the coverage area k One IRS communication; otherwise, 0. Therefore, a binary variable. The following relationship should be satisfied: (10) (11) Since this invention considers narrowband transmission, it is assumed that the reflection coefficient of the IRS is approximately constant throughout the entire signal bandwidth. This approximately flat IRS coverage area is then used to determine the IRS's reflection coefficient. k The IRS reflection matrices are denoted as , Indicates the first The phase shift vectors of IRS; where , Indicates the first The phase shift of the first reflecting element of each IRS; and so on. Indicates the first The first IRS The phase shift of each reflecting element; and , Indicates the first The phase shift of the r-th reflecting element of each IRS. For each IRS, considering the large distance between it and non-nearby BS, and the significant path loss and reflection loss experienced during signal transmission, this invention assumes that the IRS only reflects signals from nearby BS, and that non-associated IRSs are in the off state and do not reflect signals.

[0040] Since the CUAV trajectory is a preset fixed trajectory, its association sequence with the ground BS in each time slot is... This is also determined and denoted as a set. Therefore, in the nth time slot, the received signal of CUAV can be represented as the associated nth time slot. The direct connection signal of the BS, the first The superposition of the reflected signal from the IRS and co-channel interference signals from other unrelated BSs, i.e.: (12) in, The additive white Gaussian noise of the CUAV receiver has a mean of 0 and a variance of . .

[0041] Corresponding to the desired signal, Corresponding to co-channel interference signals.

[0042] Based on the above analysis, in the... The achievable signal transmission rate of a single time slot CUAV can be expressed as: (13) in, B This indicates the system communication bandwidth.

[0043] It is important to note that and In and For NLoS components, combined with the independence of fading blocks within a time slot, the time slot The transmission rate corresponding to each fading block within the block They are independent of each other, leading to It exhibits random fluctuation characteristics. Therefore, this invention focuses on the desired transmission rate, which is defined as... However, the derivation The closed-form expression for the desired transmission rate is quite challenging because its probability distribution is difficult to obtain. To address this problem, we utilize Theorem 1 to... Approximation is performed.

[0044] Theorem 1 For independent random variables X (satisfy )and Y (satisfy Approximate form Established.

[0045] make , By combining the mathematical derivation of the expectation operation with the analysis of channel characteristics, we can deduce that: (14) (15) in, , , , , , , , ; This represents the pure Loss-of-Stake (LoS) channel gain from the i-th BS to the CUAV at time slot n. This represents the pure Loss channel gain from the (b[n], k)th IRS to the CUAV at time slot n. This represents the pure Loss channel gain from the b[n]th BS to the (b[n], k)th IRS at time slot n. This represents the expected value of the squared magnitude of the composite channel. This represents the pure NLoS channel gain from the i-th BS to the CUAV at time slot n. This represents the expected value of the squared modulus of the directly connected channel. , and These represent the equivalent reference gain of the Rice channel LoS component for the BS-CUAV link, BS-IRS link, and IRS-CUAV link, respectively.

[0046] Therefore, the desired transmission rate can be derived by combining Theorem 1. The closed-form approximation expression is: (16) Meanwhile, during the CUAV task execution cycle Internally, to ensure the connectivity of the communication link between CUAV and BS, the expected transmission rate is... The following constraints also need to be met: (17) in This represents the minimum transmission rate threshold for CUAV to maintain communication connectivity. , and These respectively represent 0 IRS associated with each time slot and associated with... Transmission rate per IRS.

[0047] 4) Construct a joint optimization problem model for maximizing system EE and minimizing the number of associated IRS: during the CUAV task execution time T Within this framework, under the constraints of CUAV communication connectivity, mobility, and unit modulus, the IRS association scheduling matrix is ​​jointly optimized. BS transmit power matrix IRS phase shift matrix The aim is to achieve synergistic optimization of maximizing system EE and minimizing the number of associated IRS.

[0048] First, combining equations (9) and (16), the system energy efficiency EE can be expressed as: (18) Therefore, the corresponding multi-objective function can be initially expressed as: (19) Since the optimization objectives mentioned above differ significantly in magnitude, to avoid one objective unreasonably dominating the optimization process, a min-max normalization method is first used for preprocessing. This method maps each objective function value to the [0, 1] interval through a linear transformation, achieving scaling and uniformity of the original objective values. Specifically, let... and These respectively represent 0 IRS associated with each time slot and 1 IRS associated with each time slot. The system EE value of each IRS, then The normalized system EE can be expressed as To minimize the number of associated IRSs, by defining... Transforming this into an equivalent objective of maximizing the number of unrelated IRSs, consistent with the direction of maximizing the system's EE, the normalized result for the number of unrelated IRSs is: During the flight cycle The minimum number of internally associated IRSs is 0, and the maximum is... After normalizing the objective function, a linear weighted sum method is used to transform the multi-objective optimization problem into a single-objective optimization problem. Specifically, the system EE and the total number of unrelated IRSs are respectively determined by weighting factors. and Weighting is performed, where This is an adjustable parameter, and its magnitude directly reflects the priority of the system EE relative to the number of unrelated IRSs.

[0049] Finally, the joint optimization problem can be formulated as: (20) in, C 1 indicates the communication connection constraints when CUAV connects to BS. C 2- C 3 represents the scheduling factor constraint of the IRS. C 4 indicates that the BS's transmit power must not exceed its maximum transmit power budget. , C 5 represents the unit modulus constraint of the IRS; This represents the number of non-associative IRS after normalization; Indicates the first The first IRS r Phase shift of each reflecting element; K express A collection of IRSs deployed next to each ground-based BS; M represents the set of time slots; express A collection of ground-based BS (Browser Base Stations).

[0050] Obviously, the problem The problem is difficult to solve for the following reasons: first, the optimization variables are highly coupled, and the objective function is neither concave nor convex with respect to the optimization variables; second, the binary variables introduced by the IRS associated scheduling result in integer constraints. Therefore, the problem... It is a mixed integer non-convex optimization problem, and it is difficult to obtain the global optimal solution.

[0051] Step S4: Jointly optimize the IRS association scheduling matrix, base station transmit power matrix, and IRS phase shift matrix to achieve the optimization objective corresponding to the joint optimization problem model of maximizing system EE and minimizing the number of associated IRSs; Specifically, step S4 includes the following steps: Step S41, given the BS transmit power matrix and the IRS phase shift matrix, iteratively optimize the IRS scheduling factor matrix; Step S42, given the IRS phase shift matrix and the IRS scheduling factor matrix optimized in step S41, iteratively optimize the BS transmit power matrix; Step S43, given the IRS scheduling factor matrix and the BS transmit power matrix optimized in the above steps, iteratively optimize the IRS phase shift matrix; Step S44, determine whether the change in the optimization objectives corresponding to maximizing the system EE and minimizing the number of associated IRSs is less than a preset threshold. Or, does the current iteration count exceed the preset maximum iteration count? If any condition is met, proceed to step S5; otherwise, proceed to step S41.

[0052] Specifically, step S41 includes: Given the BS transmit power matrix IRS phase shift matrix The IRS scheduling factor matrix is ​​optimized using the initialization parameters from step S2. Then the problem (P0) can be simplified to:

[0053] ;(twenty one) To make this problem easier to handle, we will remove the binary variables in problem (P1.1). Relaxation as a continuous variable , Indicates in time slot n When the CUAV communicates with the (b(n), k)th IRS via consecutively associated scheduling factors, the optimization problem (P1) can be reformulated as:

[0054] ;(twenty two) At this point, both the objective function and constraint C1 are non-convex, because... In The components are related to the optimization variables. The component is a convex function, but when combined with the subsequent log function, the overall convexity is destroyed, making the problem difficult to solve. Linear functions, on the other hand, are compatible with log functions, ensuring the convexity of the problem. Therefore, to solve this problem, this invention transforms the component into a linear function. Specifically, an auxiliary variable is introduced. and define According to convex analysis theory, At a given local iteration point ( for No. A first-order Taylor expansion is performed at the value of the nth iteration, and the global lower bound is obtained by approximation using the SCA algorithm. The specific first-order Taylor expansion is as follows: ;(twenty three) So It can be re-represented as : ;(twenty four) in , ; This represents the cascaded equivalent channel gain of the b[n]th BS to the (b[n],k)th IRS-CUAV at time slot n; This represents a linear approximation of the squared magnitude of the channel gain. This represents the lower bound of the desired transmission rate. This represents the expected power of the NLoS component of the directly connected link. This represents the expected power of the NLoS component in the cascaded link.

[0055] Therefore, problem (P1.2) can be transformed into:

[0056] (25) At this point, problem (P1.3) is a standard convex optimization problem, which can be solved using the CVX solver to obtain a continuous solution. Subsequently, a threshold rounding strategy will be used to satisfy... Variables with a value ≥ 0.5 are set to 1, and all others are set to 0, thus restoring the binary solution. .

[0057] Specifically, step S42 includes: For a given and Problem (P0) can be simplified to:

[0058] (26) As mentioned in formula (18) ; Therefore, according to the Tinkelbach algorithm, for the optimal system EE parameters... , Equivalent to ; As mentioned in formula (16) , but Equivalent to ; in, ; ; ; In the formula, This represents the equivalent total transmission rate. This represents the logarithm of the average sum of the power of all received signals and noise during the flight time of CUAV. This represents the logarithm of the average sum of the power of the interference signal and noise received by CUAV during the flight time. Since it is the difference between two concave functions, the optimization problem still exhibits non-convexity. Given... about It is a concave function, At a given local iteration point A first-order Taylor expansion is performed at the given point, and the SCA algorithm is used for approximation to obtain its global upper bound. This indicates that at time slot n, the first... The BS transmit power value is obtained in the next iteration. The specific first-order Taylor expansion is as follows: (27) in express A linear approximation.

[0059] Therefore, we can obtain An approximate lower bound: (28) Therefore, the optimization problem can be reformulated as follows:

[0060] (29) Clearly, question (P2.2) is relevant. It is convex and can be solved using the CVX solver.

[0061] It should be noted that solving problem (P2.2) is an iterative optimization process, and The value needs to be updated in each iteration. Initialized From initialization , and The calculation is obtained, and then the calculation is obtained. right , and Perform iterative optimization.

[0062] Specifically, step S43 includes: For a given IRS scheduling factor matrix and BS transmit power matrix Problem (P0) can be simplified to:

[0063] (30) Given the IRS scheduling factor matrix obtained in the previous optimization step and BS transmit power matrix The IRS phase shift matrix is ​​optimized using the initialization parameters obtained in step S2. For this non-convex optimization subproblem, we can introduce auxiliary variables... and The SCA algorithm and SDR technology were used to process and obtain information about The standard convex optimization problem, namely:

[0064] (31) in, This represents the sum of all elements on the main diagonal of the matrix; The autocorrelation matrix represents the equivalent complex gain matrix of the CUAV received signal. express The autocorrelation matrix of the augmented vector; , , This represents the equivalent complex gain matrix of the CUAV received signal. express The augmented vector; where, This represents the total number of IRS associated with CUAV under the b[n]th BS at time slot n; , , ; ; ; This represents the "preprocessing matrix" of the equivalent concatenated channel from BS to CUAV. This represents the matrix formed by combining the phase shift vectors of the associated IRS within the b[n]th BS under time slot n; This represents the set of IRS associated sequences communicating with CUAV within the range of the b[n]th associated BS at time slot n. Must meet and Since the rank-one constraint is non-convex, we use the SDR technique to relax it. Indicates the nth time slot The Each element.

[0065] Problem (P3.2) is a standard SDP problem, solvable using existing tools like CVX. Generally, problem (P3.2) cannot be solved... The optimal solution is rank one, which means that the optimal objective value of the problem can only serve as the upper bound of problem (P3.1); therefore, some additional steps are needed to further construct the optimal high-rank solution of (P3.2) into a rank one solution.

[0066] In this invention, we use the Gaussian randomization method to solve this problem. The specific steps are as follows: First, for Perform eigenvalue decomposition to obtain ,in , They are unitary and diagonal matrices, respectively, both having dimensions of 1. Subsequently, the suboptimal solution to problem (P3.1) can be obtained as follows: ,in It is based on The generated random vector, here This indicates that the mean is 0 and the covariance matrix is ​​the identity matrix. The CSCG distribution is determined by iterating through all independently generated Gaussian random vectors. Select the option that maximizes the objective function value in problem (P3.2). As an approximate solution. Finally, through The IRS phase shift in the recovery problem (P3.1) is addressed, where... Representing vectors The former Each element.

[0067] Specifically, step S44 includes: Based on the output of step S43, determine whether the changes in the optimization objectives corresponding to minimizing the system EE and the number of associated IRSs are both less than the preset threshold. Or, does the current iteration count exceed the preset maximum iteration count? If any condition is met, proceed to step S5; otherwise, proceed to step S41.

[0068] Step S5: Output the optimal trade-off between maximizing the system EE and minimizing the number of associated IRSs; After iterative optimization, the optimal trade-off between the number of system EEs and associated IRSs is obtained, and the results are output, thus concluding the process.

[0069] To verify the effectiveness of the proposed IRS-aware, networked UAV high-energy-efficiency communication multi-dimensional resource optimization method, MATLAB was used to conduct simulation experiments on the proposed multi-dimensional resource optimization method. Considering a three-dimensional Cartesian coordinate system, the deployment... There are 10 BS (Base Stations), all of which are evenly distributed throughout the coverage area, with a height of 1000 meters. Each BS is equipped with The antenna provides a base station (BS) coverage radius of 100m. Around each BS, six IRSs are evenly deployed at a radius of 5m, all at the same height. The horizontal coordinates of the start and end points of CUAV are respectively... , Flight altitude Time slot length That is, each time slot The duration of n is 1 second. , The path loss exponents for BS-UAV, BS-IRS, and IRS-UAV are respectively , Rice factor The antenna spacing between the BS and IRS meets the following requirements. Additive white Gaussian noise power Communication bandwidth Convergence threshold By default, the weighting factors are all SE threshold .

[0070] To evaluate the performance advantages of the proposed IRS-aware, multi-dimensional resource optimization method for high-energy-efficiency communication of connected UAVs, five typical benchmark schemes were selected for performance comparison: Benchmark Scheme 1, which optimizes only the BS transmit power and IRS associated scheduling, with IRS phase shifts randomly generated; Benchmark Scheme 2, which optimizes BS transmit power and IRS phase shifts but not IRS scheduling; Benchmark Scheme 3, which optimizes only the BS transmit power but not IRS associated scheduling and phase shifts; Benchmark Scheme 4, which empowers the CUAV communication system with a single IRS, i.e., only one IRS is deployed next to each BS, and all variables are jointly optimized; Benchmark Scheme 5: "No IRS", which allows... That is, to maximize the system EE without IRS enablement.

[0071] Figure 3As can be seen, with the increase of the number of iterations, the system EE of all schemes gradually increases and tends to stabilize after about 5 iterations. Furthermore, compared with all benchmark schemes, the proposed scheme has the highest system EE, proving that jointly optimizing IRS scheduling and phase shift can significantly improve system EE. It also shows that the proposed scheme delivers a more significant EE gain than the scheme of deploying a single IRS next to the BS. In addition, compared with benchmark scheme 1, benchmark scheme 2 achieves a higher system EE. This is because the core of IRS phase shift optimization is to reconstruct the wireless channel to enhance the strength of the desired signal and suppress interference. IRS scheduling, as a resource allocation optimization, has limited effect on enhancing the desired signal. Therefore, IRS-optimized phase shift has a better performance gain than IRS scheduling optimization in improving system EE.

[0072] from Figure 4 It can be observed that the system EE and the number of associated IRS for all schemes tend to stabilize around the fifth iteration. The system EE increases rapidly in the early stages of iteration, while the number of associated IRS decreases rapidly in the early stages of iteration. Furthermore, with the increase of CUAV flight time and weighting factor, both the system EE and the number of associated IRS for the proposed schemes show an upward trend. It is worth noting that increasing the weighting factor can effectively alleviate the problem of decreased system EE caused by the shortening of CUAV flight time. This is because the weighting factor... The study directly adjusted the trade-off between system EE and the number of associated IRS, increasing the weighting factor to prioritize maximizing system EE and thus improving it. In contrast, while extending CUAV flight time also helps improve system EE, it increases CUAV propulsion energy consumption, resulting in a relatively limited improvement. This result demonstrates that a reasonable allocation of weighting factors can achieve an effective trade-off between different objectives, thereby effectively improving system EE.

[0073] from Figure 5 and Figure 6 It can be seen that the number of system EEs and associated IRSs in all schemes increases with... The number of non-related IRSs increases with the increase of [unrelated information]. The increase in the number of associated IRSs leads to a decrease in the efficiency. This result confirms the conflict between maximizing the system energy efficiency (EE) and minimizing the number of associated IRSs. Compared to the baseline scheme 1, the proposed scheme improves the system EE but also associates more IRSs. This is because the baseline scheme 1 does not perform passive beamforming optimization on the IRSs, resulting in the inability to coherently superimpose multiple link signals at the receiver, thus limiting the system EE. To balance maximizing the system EE with minimizing the number of associated IRSs, this scheme actively eliminates IRSs that do not contribute to the system energy efficiency or even have a negative impact. This trade-off mechanism reduces the number of associated IRSs but also sacrifices the system EE, making it inferior to the proposed scheme. Furthermore, comparisons at different flight times further show that extending the flight time of the CUAV can simultaneously improve both the system EE and the number of associated IRSs. This phenomenon verifies the positive gain effect of the CUAV flight time on system performance.

[0074] Figure 7 As can be seen, all curves exhibit a fluctuating trend due to the periodic changes in the channel caused by CUAV mobility. The proposed scheme outperforms the baseline scheme in overall performance. It is noteworthy that the SE of the baseline scheme 4 and the scheme without IRS repeatedly falls below the preset SE threshold, facing the risk of communication interruption; while the proposed scheme ensures that the SE is above this threshold in all time slots, and further... As the SE curve increases, the baseline scheme 4 relies solely on a single IRS near the BS. When the CUAV flies away from the BS to the edge region, a single IRS cannot provide sufficient reflection gain to overcome the path loss caused by long-distance transmission. In contrast, the proposed scheme dynamically schedules multiple IRSs near the BS, selecting the IRS with the best communication link for auxiliary communication, effectively overcoming the coverage blind spots of a single IRS. This result demonstrates that a multi-IRS collaborative strategy can effectively improve SE, thereby ensuring the user's communication connectivity requirements throughout the entire time period.

Claims

1. A multi-dimensional resource optimization method for networked unmanned aerial vehicle (UAV) communication based on IRS quantity awareness, characterized in that, Specifically, the following steps are included: Step S1, Method begins; Step S2: Initialize base station information, CUAV information, and IRS information; wherein, the base station information includes the number of base stations deployed, altitude, coordinates, number of antennas equipped, and number of IRS deployed next to each base station; the CUAV information includes the CUAV flight trajectory; and the IRS information includes the IRS deployment altitude, coordinates, and number of reflective units equipped. Step S3: Based on base station information, CUAV information, and IRS information, establish a system channel model, energy consumption model, and IRS quantity-aware CUAV downlink communication model; based on the system channel model, energy consumption model, and IRS quantity-aware CUAV downlink communication model, establish a joint optimization problem model for maximizing system EE and minimizing associated IRS quantity. Step S4: Jointly optimize the IRS association scheduling matrix, base station transmit power matrix, and IRS phase shift matrix to achieve the optimization objective corresponding to the joint optimization problem model of maximizing system EE and minimizing the number of associated IRSs; Step S5: Output the optimal compromise value of the optimization objective corresponding to the joint optimization problem model of maximizing the system EE and minimizing the number of associated IRS.

2. The IRS-aware multi-dimensional resource optimization method for networked UAV communication according to claim 1, characterized in that: Step S4 specifically includes: Step S41: Given the BS transmit power matrix and the IRS phase shift matrix, iteratively optimize the IRS scheduling factor matrix; Step S42: Given the IRS phase shift matrix and the IRS scheduling factor matrix obtained in step S41, iteratively optimize the BS transmit power matrix; Step S43: Given the IRS scheduling factor matrix and BS transmit power matrix obtained from the above steps, iteratively optimize the IRS phase shift matrix; Step S44: Determine whether the changes in the optimization objectives corresponding to maximizing the system EE and minimizing the number of associated IRSs are less than a preset threshold. Or, does the current iteration count exceed the preset maximum iteration count? If any condition is met, proceed to step S5; otherwise, proceed to step S41.

3. The IRS-aware multi-dimensional resource optimization method for networked UAV communication according to claim 2, characterized in that, The step S3, which establishes the system channel model based on base station information, CUAV information, and IRS information, specifically includes: Acquiring time slots n , No. m Between the first BS and CUAV, the first Between the IRS and CUAV and the first m The BS and the first Channel gain between IRS , and , This represents the index of the k-th IRS next to the m-th BS; This indicates the number of antennas equipped in each BS. The number of reflective units equipped in the IRS is indicated by the following formula: ； ； ； , and They represent time slots respectively. n No. m The three-dimensional distance of the first BS-CUAV, the first The three-dimensional distance of the first IRS-CUAV and the first m The BS-then Three-dimensional distance of each IRS; This represents the channel gain at a reference distance of 1m. , and These represent the path loss factors for the BS-CUAV link, IRS-CUAV link, and BS-IRS link, respectively. , and These represent the Rice factors for the BS-CUAV link, IRS-CUAV link, and BS-IRS link, respectively. , and They represent time slots respectively n , No. m Between the first BS and CUAV, the first Between the IRS and CUAV and the first m The BS and the first The deterministic LoS component of the channel gain between IRSs, and , and This corresponds to the NLoS component that follows a Rayleigh distribution; in, and The calculation formula is: ； ； in The working carrier wavelength, and These represent the BS antenna spacing and the IRS element spacing, respectively; and and They represent time slots respectively n , No. m The pitch departure angle and azimuth departure angle of the downlink from BS to CUAV satisfy the following relationship: ; and They represent time slots respectively n At that time, the first The pitch departure angle and azimuth departure angle of the downlink from the IRS to CUAV satisfy the following conditions: as well as Among them, the first m The BS and the first The horizontal coordinates of the IRS are respectively ;CUAV in the n The horizontal coordinate of the time slot is represented as ; Indicates the height of CUAV. Indicates the height of the IRS; at the same time, The calculation formula is: ； ； in, and They represent the first m BS to the first The pitch and azimuth angles of arrival of each IRS downlink, and , , ;in, Indicates the first m The BS and the first The array response vector at the IRS end between each IRS; Then it means the first m The BS and the first Between each IRS, the array response vector at the BS end, Indicates the height of BS.

4. The IRS-aware multi-dimensional resource optimization method for networked UAV communication according to claim 3, characterized in that, In step S3, an energy consumption model is specifically established based on base station information, CUAV information, and IRS information. include: Among them, during the flight cycle The total energy consumption of the internal system is expressed as: ； Indicates time slot n No. m The transmit power of each BS; Indicates the time slot length; This indicates that CUAV is in flight cycle T The propulsion power consumption within the system is calculated using the following formula: ; In the formula, Indicates in time slot At that time, the flight speed of CUAV; and These represent the blade power and induced power under CUAV hovering conditions, respectively. , ; The rotor blade tip velocity, The average rotor induced velocity during hovering. and These represent air density and rotor disk area, respectively. and This indicates the fuselage drag ratio and rotor rigidity; Indicates the angular velocity of the blade; Indicates the rotor radius; Indicates the weight of the CUAV; Indicates the airfoil drag coefficient; This represents the incremental correction coefficient for induced power.

5. The IRS-aware multi-dimensional resource optimization method for networked UAV communication according to claim 4, characterized in that, The step S3, which establishes the IRS quantity-aware CUAV downlink communication model based on base station information, CUAV information, and IRS information, specifically includes: Obtain CUAV in the The achievable signal transmission rate for each time slot is calculated using the following formula: ； In the formula, Indicates the first Time slot, CUAV associated BS; B Indicates system communication bandwidth. Represents the square of the Euclidean norm; Indicates conjugate transpose; , and They represent time slots respectively n , No. Between the first BS and CUAV, the first Between the IRS and CUAV and the first The BS and the first Channel gain between IRS; Indicates CUAV and the first A binary scheduling factor associated with each IRS; Indicates the first The k-th IRS reflection matrix within the coverage area of ​​each base station; Indicates time slot n No. The transmit power of each BS; Indicates in time slot n The channel gain between the i-th BS and CUAV; Time slot n The transmit power of the i-th BS; M express The collection of ground-based BS; This represents the power of additive white Gaussian noise; For CUAV in the The expected achievable signal transmission rate of each time slot is obtained by calculating the following formula: ； in, ; ； In the formula, , , , , , , , ; This represents the pure Loss-of-Stake (LoS) channel gain from the i-th BS to the CUAV at time slot n. This represents the pure Loss channel gain from the (b[n], k)th IRS to the CUAV at time slot n. This represents the pure Loss channel gain from the b[n]th BS to the (b[n], k)th IRS at time slot n. This represents the expected value of the squared magnitude of the composite channel. This represents the pure NLoS channel gain from the i-th BS to the CUAV at time slot n. This represents the expected value of the squared modulus of the directly connected channel. , and These represent the equivalent reference gain of the Ricean channel LoS component for the BS-CUAV link, BS-IRS link, and IRS-CUAV link, respectively. , and They represent time slots respectively n No. The three-dimensional distance of the first BS-CUAV, the first The three-dimensional distance of the first IRS-CUAV and the first The BS-then The three-dimensional distance of each IRS.

6. The IRS-aware multi-dimensional resource optimization method for networked unmanned aerial vehicle (UAV) communication according to claim 5, characterized in that, Step S3, which establishes a joint optimization problem model for maximizing system EE and minimizing associated IRS based on the system channel model, energy consumption model, and IRS number-aware CUAV downlink communication model, specifically includes: Construct the objective function to maximize the system's energy expenditure (EE) and the objective function to minimize the number of associated IRSs: ； Wherein, system EE can be represented as ; X represents the IRS associated scheduling matrix. Represents the BS transmit power matrix; Represents the IRS phase shift matrix; After normalizing the objective function, the multi-objective optimization problem is transformed into a single-objective optimization problem using a linear weighted sum method, resulting in a joint optimization problem model that maximizes the system's energy efficiency (EE) and minimizes the number of associated information sources (IRS). ； in, C 1 indicates the communication connection constraints when CUAV connects to BS. C 2- C 3 represents the scheduling factor constraint of the IRS. C 4 indicates that the BS's transmit power must not exceed its maximum transmit power budget. , C 5 represents the unit modulus constraint of the IRS; This represents the minimum transmission rate threshold for CUAV to maintain communication connectivity. It is an adjustable parameter. This represents the normalized system EE. This represents the number of non-associative IRS after normalization; Indicates the first The first IRS r Phase shift of each reflecting element; K express A collection of IRSs deployed next to each ground-based BS; Represents a set of time slots.

7. The IRS-aware multi-dimensional resource optimization method for networked unmanned aerial vehicle communication according to claim 6, characterized in that, Step S41 specifically includes: Given the BS transmit power matrix IRS phase shift matrix The joint optimization problem of maximizing system EE and minimizing the number of associated IRSs is simplified to: ； The binary variables in problem (P1.1) Relaxation as a continuous variable , Indicates in time slot n When the scheduling factors of CUAV and the (b(n), k)th IRS communication are continuously associated, the model is reformulated as follows: ； Introducing auxiliary variables and define ,Will In Components at a given local iteration point A first-order Taylor expansion is performed at the given location, and the SCA algorithm is used for approximation to obtain its global lower bound; where, for No. The value of the nth iteration; This represents the cascaded equivalent channel gain of the b[n]th BS to the (b[n],k)th IRS-CUAV at time slot n; but Represented as , Indicates the lower bound of the desired transmission rate: ； in , ; ； This represents the expected power of the NLoS component of the directly connected link; This represents the expected power of the NLoS component in the cascaded link; This represents a linear approximation of the squared magnitude of the channel gain; The model then transforms into: ； In the formula, This represents the total number of time slots. express The total number of IRS deployed next to each ground-based BS; and These respectively represent 0 IRS associated with each time slot and 1 IRS associated with each time slot. The system EE value of each IRS.

8. The IRS-aware multi-dimensional resource optimization method for networked unmanned aerial vehicle communication according to claim 6, characterized in that, Step S42 specifically includes: Given and The problem (P0) can be simplified to: ； According to Tinkelbach's algorithm, Transform into ; in, , For optimal system EE parameters; ; ; In the formula, This represents the equivalent total transmission rate. This represents the logarithm of the average sum of the power of all received signals and noise during the flight time of CUAV; This represents the logarithm of the average sum of the power of the interference signals and noise received by CUAV during the flight time. Will At a given local iteration point A first-order Taylor expansion is performed at the given point, and the SCA algorithm is used for approximation to obtain its global upper bound. This indicates that at time slot n, the first... The BS transmit power value is obtained in the next iteration; the specific first-order Taylor expansion is: ； in, express A linear approximation of yields... An approximate lower bound: ； The model then transforms into: 。 9. A multi-dimensional resource optimization method for networked UAV communication based on IRS quantity awareness according to claim 6, characterized in that, Step S43 specifically includes: For a given IRS scheduling factor matrix and BS transmit power matrix Problem (P0) simplifies to: ； Introducing auxiliary variables and The SCA algorithm and SDR technology were used to process and obtain information about The standard convex optimization problem is then expressed by the model as follows: ； in, This represents the sum of all elements on the main diagonal of the matrix; The autocorrelation matrix represents the equivalent complex gain matrix of the CUAV received signal. express The autocorrelation matrix of the augmented vector; , ;in, This represents the total number of IRS associated with CUAV under the b[n]th BS at time slot n; , , This represents the equivalent complex gain matrix of the CUAV received signal. express augmented vector; ； ； ； This represents the equivalent concatenated channel preprocessing matrix from BS to CUAV. This represents the matrix formed by combining the phase shift vectors of the associated IRS within the b[n]th BS under time slot n; This represents the set of IRS associated sequences communicating with CUAV within the range of the b[n]th associated BS at time slot n; Must meet and ; Indicates the nth time slot The Each element.

Images