Resource allocation method and system in space-air-ground integrated Internet of Things data acquisition system

Abstract

The invention discloses a resource allocation method in a space-air-ground integrated Internet of Things data acquisition system. The method comprises the following steps: constructing a wireless communication model of the space-air-ground integrated Internet of Things data acquisition system; representing the throughput of a ground-air access link and an air-air return link in the data acquisition process; constructing a resource allocation model in the space-air-ground integrated Internet of Things data acquisition system by using an optimization problem P1 of maximizing the minimum throughput of all ground-air access links; introducing a non-negative auxiliary variable, and equivalently converting the optimization problem P1 into an optimization problem P2; and decomposing the optimization problem P2 into sub-problems, and solving the resource allocation model in the space-air-ground integrated Internet of Things data acquisition system according to a preset efficient algorithm. According to the method, a complex communication scene of multiple devices, multiple unmanned aerial vehicles and multiple low-orbit satellites is considered, and under a more universal probability channel model, three-dimensional node pairing, bandwidth allocation and UAV base station layout multi-dimensional communication resources are jointly optimized, the minimum throughput of all data acquisition links is maximized, and the data acquisition efficiency is improved. Therefore, the user fairness of the whole system is improved in a high-density communication scene.

Description

Technical Field

[0001] The present invention relates to a resource allocation method and system in an air-ground-integrated Internet of Things data acquisition system, belonging to the technical field of wireless communications. Background Art

[0002] The integrated space-air-ground network is an emerging network architecture that aims to achieve full coverage and seamless communication capabilities by integrating satellite network layers, air network layers and ground network layers. Among them, the satellite network layer is mainly composed of low earth orbit (LEO) satellites, which are responsible for providing wide-area coverage and long-distance communication support; the air network layer is mainly composed of unmanned aerial vehicles (UAVs), which serve as flexible relay nodes or service providers to enhance regional communication capabilities and network scalability; the ground network layer is composed of ground Internet of Things (IoT) devices, covering fixed base stations, mobile terminals and other communication facilities, responsible for providing local communication services and user access.

[0003] Due to the advantages of seamless coverage and high data transmission rate of the integrated space-ground network, the integrated space-ground network has become one of the emerging technologies of future wireless communication systems. In recent years, the integrated space-ground network has gradually become a research hotspot for many major projects around the world, such as the Global Information Grid, the OneWeb plan, and SpaceX's Starlink plan. With its advantages of wide coverage, high throughput, and strong flexibility, the integrated space-ground network has shown great potential in many fields, including earth observation, military missions, disaster relief, and remote communications.

[0004] However, compared with the traditional single-structure network, the integrated space-ground network, as a highly heterogeneous network, is limited by the limited and unbalanced communication resources in the three network layers. How to improve the system throughput under limited network resources while ensuring fairness among users has become the core problem of current research. Therefore, research on bandwidth resource allocation, dynamic device management, and user fairness has important academic value and practical significance for promoting the sustainable development of the integrated space-ground network.

[0005] In view of the above background, the present invention is proposed. Summary of the invention

[0006] The present invention provides a resource allocation method in an air-space-ground-integrated Internet of Things data acquisition system, aiming to optimize resource allocation, improve the throughput and user fairness of the air-space-ground-integrated Internet of Things system, and provide an effective technical means for solving the problem of resource allocation in heterogeneous networks.

[0007] The technical solution of the present invention is:

[0008] According to a first aspect of the present invention, a method for allocating resources in an air-ground-integrated Internet of Things data acquisition system is provided, comprising the following steps:

[0009] S1. Construct a wireless communication model of an air-ground-integrated IoT data collection system; the wireless communication model for constructing an air-ground-integrated IoT data collection system includes K LEO satellites, N multi-rotor UAV base stations, and M IoT devices, a ground-to-air access link is established between the IoT device and the UAV base station, and an air-to-space return link is established between the UAV base station and the LEO satellite;

[0010] S2, characterizes the throughput of the ground-to-air access link and the air-to-space backhaul link during data collection;

[0011] S3, constructing a resource allocation model in the air-ground integrated IoT data collection system based on the optimization problem P1 of maximizing the minimum throughput of all ground-air access links;

[0012] S4. Introduce non-negative auxiliary variables to convert the optimization problem P1 into the optimization problem P2; decompose the optimization problem P2 into sub-problems, and solve the resource allocation model in the integrated air-space-ground Internet of Things data acquisition system based on the preset efficient algorithm.

[0013] Furthermore, the throughput of the ground-to-air access link and the air-to-space backhaul link during the characterization data collection process includes:

[0014] According to the positional relationship between UAV base stations and IoT devices, a generalized probability propagation model is used to calculate the channel gain between each UAV base station and IoT device while considering both line-of-sight and non-line-of-sight propagation channels. Based on the channel gain between each UAV base station and IoT device, the throughput of the ground-to-air access link during data collection is calculated in combination with the Shannon formula.

[0015] According to the positional relationship between LEO satellites and UAV base stations, combined with the Rician fast fading model, the channel gain between each LEO satellite and the UAV base station is calculated; according to the channel gain between each LEO satellite and the UAV base station, combined with the Shannon formula, the throughput of the space-to-space backhaul link during the data acquisition process is calculated.

[0016] Furthermore, the construction of a resource allocation model in the air-space-ground-integrated Internet of Things data acquisition system specifically comprises: in the air-space-ground-integrated Internet of Things data acquisition system, by jointly optimizing the multi-dimensional communication resources of three-dimensional node pairing, bandwidth allocation and UAV base station layout, an objective function and constraints are established for the optimization problem P1 of maximizing the minimum throughput of all ground-to-air access links to construct a resource allocation model in the air-space-ground-integrated Internet of Things data acquisition system.

[0017] Furthermore, the resource allocation model in the air-ground-integrated IoT data acquisition system is expressed as:

[0018]

[0019] Where, the goal of the optimization problem P1 is to maximize the minimum throughput of all ground-to-air access links; constraint C1 represents the bandwidth allocation variable of IoT devices and UAV base stations is between 0 and 1; Constraint C2 requires that the sum of IoT device bandwidth allocation variables and the sum of UAV base station bandwidth allocation variables cannot exceed 1; Constraint C3 requires that the IoT device attachment optimization variable used for 3D node pairing Drone attachment optimization variables All are binary values; Constraint C4 requires that each IoT device can only connect to one UAV base station at most, and each UAV base station can only connect to one LEO satellite; Constraint C5 requires that the throughput of the ground-to-air access link for any UAV base station is less than or equal to the throughput of the air-to-space return link; Constraint C6 requires that the flight speed of the UAV base station Less than or equal to ζ max ; Constraints C7-C9 require that the horizontal coordinates of each UAV base station are within a circular area with a radius of r, and the flight altitude has an upper limit z max and the lower limit z min , and the distance between any two UAV base stations is not less than the safety distance d min , Constraint C10 requires that the energy consumption of each UAV base station be capped at Q max , Indicates the initial position of all UAV base stations; the spatial position of the nth UAV base station Q(0) represents the energy consumption of each UAV in collecting data when it is stationary in the hovering position; Indicates that the UAV base station flying speed is The energy consumption of collecting data.

[0020] Furthermore, the optimization problem P2 is decomposed into sub-problems, specifically including: a three-dimensional node pairing sub-problem, a bandwidth allocation sub-problem, and a UAV base station layout sub-problem.

[0021] Furthermore, solving the resource allocation model in the air-ground-integrated IoT data collection system according to a preset efficient algorithm includes:

[0022] S4.1. When the bandwidth allocation and UAV base station layout variables are given, the remaining three-dimensional node pairing subproblems can be constructed as optimization problem P3, which is solved by linear integer programming tools;

[0023] S4.2, when the three-dimensional node pairing and UAV base station layout are given, the remaining bandwidth allocation sub-problem can be constructed as optimization problem P4, and the optimization problem P4 is written as optimization problem P5 using a continuous convex approximation algorithm for optimization solution;

[0024] S4.3, when the three-dimensional node pairing and bandwidth allocation variables are given, the remaining UAV base station layout subproblems are constructed as optimization problem P6, which is solved using the sequential quadratic programming algorithm;

[0025] S4.4. Combine S4.1-S4.3 above and use the alternating optimization method to obtain the resource allocation result of the integrated air-space-ground-integrated Internet of Things data acquisition system.

[0026] Furthermore, the S4.4 is specifically:

[0027] Initialization: Set t=1 to the current number of iterations, T max Set to the maximum number of iterations; given the initial variables that satisfy the constraints of the optimization problem P2 After initialization, perform the following steps:

[0028] Step 1: In the current optimization variable Under these conditions, the suboptimal solution of the three-dimensional node pairing subproblem is obtained through S4.1 optimization and updated to

[0029] Step 2: In the current optimization variable Under these conditions, the suboptimal solution of the current bandwidth allocation subproblem is obtained through S4.2 optimization and updated to

[0030] Step 3: In the current optimization variable Under these conditions, the suboptimal solution of the UAV base station layout subproblem is obtained through S4.3 optimization and updated to

[0031] Step 4: If t ≥ 2 and θ (t) -θ (t-1) ≤ε 2 , then the proposed algorithm converges, and the current This is the suboptimal solution to the optimization problem P1; otherwise, set t=t+1 and return to Step 1.

[0032] According to a second aspect of the present invention, a resource allocation system in an air-space-ground-integrated Internet of Things data collection system is provided, comprising a module of a resource allocation method in an air-space-ground-integrated Internet of Things data collection system as described in any one of the above.

[0033] According to a third aspect of the present invention, a processor is provided, wherein the processor is used to run a program, and when the program is run, the resource allocation method in the air-ground-integrated Internet of Things data acquisition system described in any one of the above is executed.

[0034] The beneficial effects of the present invention are:

[0035] First, the present invention takes into account the complex communication scenarios of multiple devices, multiple drones, and multiple low-orbit satellites, and jointly optimizes three-dimensional node pairing, bandwidth allocation, and UAV base station layout multi-dimensional communication resources under a more universal probabilistic channel model to maximize the minimum throughput of all data acquisition links, thereby improving user fairness of the entire system in high-density communication scenarios.

[0036] Second, for the optimization problem P1, in order to reduce the computational complexity and improve the practicality of the algorithm, the present invention cleverly decomposes the optimization problem P1 into three sub-problems of three-dimensional node pairing, bandwidth allocation and UAV base station layout through decoupling; on the basis of the above, the three-dimensional node pairing sub-problem can be directly solved by linear integer programming tools, and based on this solution method, the solution of the sub-problem can reach the global optimum; based on the characteristic that the bandwidth allocation sub-problem is easy to approximate the convex optimization problem, the present invention proposes an optimization algorithm based on the continuous convex approximation theory, which can further achieve the continuous optimization of the approximated convex optimization problem on the basis of approximating the original sub-problem, so that the solution of the sub-problem is better; considering the non-convexity and nonlinear characteristics of the UAV base station layout sub-problem, the present invention adopts an optimization algorithm based on the sequential quadratic programming theory, based on which the sub-optimal solution of the sub-problem can be effectively obtained; on this basis, in order to avoid falling into the local optimal solution, the above sub-problems are further jointly optimized through the alternating optimization theory, so as to converge to a sub-optimal solution of the original problem. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 It is a flowchart of the method of the present invention;

[0038] Figure 2 A wireless communication model of an air-ground-integrated Internet of Things data acquisition system provided according to an embodiment of the present invention;

[0039] Figure 3 This is the convergence analysis of the algorithm proposed in the present invention;

[0040] Figure 4 The minimum throughput of the data acquisition link under different numbers of UAV base stations is provided to demonstrate various mechanisms according to an embodiment of the present invention. DETAILED DESCRIPTION

[0041] In order to make the purpose, technical scheme and advantages of the embodiments of the present invention clearer, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention. It should be noted that the embodiments in this application and the features in the embodiments can be combined with each other arbitrarily without conflict.

[0042] Example 1: Figure 1-Figure 2 As shown, according to a first aspect of an embodiment of the present invention, a resource allocation method in an air-ground-integrated Internet of Things data acquisition system is provided, comprising the following steps:

[0043] S1. Construct a wireless communication model of an air-ground-integrated IoT data collection system; the wireless communication model for constructing an air-ground-integrated IoT data collection system includes K LEO satellites, N multi-rotor UAV base stations, and M IoT devices, a ground-to-air access link is established between the IoT device and the UAV base station, and an air-to-space return link is established between the UAV base station and the LEO satellite;

[0044] S2, characterizes the throughput of the ground-to-air access link and the air-to-space backhaul link during data collection;

[0045] S3, constructing a resource allocation model in the air-ground integrated IoT data collection system based on the optimization problem P1 of maximizing the minimum throughput of all ground-air access links;

[0046] S4. Introduce non-negative auxiliary variables to convert the optimization problem P1 into the optimization problem P2; decompose the optimization problem P2 into sub-problems, and solve the resource allocation model in the integrated air-space-ground Internet of Things data acquisition system based on the preset efficient algorithm.

[0047] Furthermore, the throughput of the ground-to-air access link and the air-to-space backhaul link during the characterization data collection process includes:

[0048] According to the positional relationship between UAV base stations and IoT devices, a generalized probability propagation model is used to calculate the channel gain between each UAV base station and IoT device while considering both line-of-sight and non-line-of-sight propagation channels. Based on the channel gain between each UAV base station and IoT device, the throughput of the ground-to-air access link during data collection is calculated in combination with the Shannon formula.

[0049] According to the positional relationship between LEO satellites and UAV base stations, combined with the Rician fast fading model, the channel gain between each LEO satellite and the UAV base station is calculated; according to the channel gain between each LEO satellite and the UAV base station, combined with the Shannon formula, the throughput of the space-to-space backhaul link during the data acquisition process is calculated.

[0050] Furthermore, the construction of a resource allocation model in the air-space-ground-integrated Internet of Things data acquisition system specifically comprises: in the air-space-ground-integrated Internet of Things data acquisition system, by jointly optimizing the multi-dimensional communication resources of three-dimensional node pairing, bandwidth allocation and UAV base station layout, an objective function and constraints are established for the optimization problem P1 of maximizing the minimum throughput of all ground-to-air access links to construct a resource allocation model in the air-space-ground-integrated Internet of Things data acquisition system.

[0051] Furthermore, the resource allocation model in the air-ground-integrated IoT data acquisition system is expressed as:

[0052]

[0053] Where, the goal of the optimization problem P1 is to maximize the minimum throughput of all ground-to-air access links; constraint C1 represents the bandwidth allocation variable of IoT devices and UAV base stations is between 0 and 1; Constraint C2 requires that the sum of IoT device bandwidth allocation variables and the sum of UAV base station bandwidth allocation variables cannot exceed 1; Constraint C3 requires that the IoT device attachment optimization variable used for 3D node pairing Drone attachment optimization variables All are binary values; Constraint C4 requires that each IoT device can only connect to one UAV base station at most, and each UAV base station can only connect to one LEO satellite; Constraint C5 requires that the throughput of the ground-to-air access link for any UAV base station is less than or equal to the throughput of the air-to-space return link; Constraint C6 requires that the flight speed of the UAV base station Less than or equal to ζ max ; Constraints C7-C9 require that the horizontal coordinates of each UAV base station are within a circular area with a radius of r, and the flight altitude has an upper limit z max and the lower limit z min , and the distance between any two UAV base stations is not less than the safety distance d min , Constraint C10 requires that the energy consumption of each UAV base station be capped at Q max , Indicates the initial position of all UAV base stations; the spatial position of the nth UAV base station Q(0) represents the energy consumption of each UAV in collecting data when it is stationary in the hovering position; Indicates that the UAV base station flying speed is The energy consumption of collecting data.

[0054] Furthermore, the optimization problem P2 is decomposed into sub-problems, specifically including: a three-dimensional node pairing sub-problem, a bandwidth allocation sub-problem, and a UAV base station layout sub-problem.

[0055] Furthermore, solving the resource allocation model in the air-ground-integrated IoT data collection system according to a preset efficient algorithm includes:

[0056] S4.1. When the bandwidth allocation and UAV base station layout variables are given, the remaining three-dimensional node pairing subproblems can be constructed as optimization problem P3, which is solved by linear integer programming tools;

[0057] S4.2, when the three-dimensional node pairing and UAV base station layout are given, the remaining bandwidth allocation sub-problem can be constructed as optimization problem P4, and the optimization problem P4 is written as optimization problem P5 using a continuous convex approximation algorithm for optimization solution;

[0058] S4.3, when the three-dimensional node pairing and bandwidth allocation variables are given, the remaining UAV base station layout subproblems are constructed as optimization problem P6, which is solved using the sequential quadratic programming algorithm;

[0059] S4.4. Combine S4.1-S4.3 above and use the alternating optimization method to obtain the resource allocation result of the integrated air-space-ground-integrated Internet of Things data acquisition system.

[0060] Furthermore, the S4.4 is specifically:

[0061] Initialization: Set t=1 to the current number of iterations, T max Set to the maximum number of iterations; given the initial variables that satisfy the constraints of the optimization problem P2 After initialization, perform the following steps:

[0062] Step 1: In the current optimization variable Under these conditions, the suboptimal solution of the three-dimensional node pairing subproblem is obtained through S4.1 optimization and updated to

[0063] Step 2: In the current optimization variable Under these conditions, the suboptimal solution of the current bandwidth allocation subproblem is obtained through S4.2 optimization and updated to

[0064] Step 3: In the current optimization variable Under these conditions, the suboptimal solution of the UAV base station layout subproblem is obtained through S4.3 optimization and updated to

[0065] Step 4: If t ≥ 2 and θ (t) -θ (t-1) ≤ε 2 , then the proposed algorithm converges, and the current This is the suboptimal solution to the optimization problem P1; otherwise, set t=t+1 and return to Step 1.

[0066] According to the second aspect of an embodiment of the present invention, a resource allocation system in an air-space-ground integrated Internet of Things data acquisition system is provided, including a module of a resource allocation method in an air-space-ground integrated Internet of Things data acquisition system described in any one of the above. Specifically comprising: a first module, for executing S1, constructing a wireless communication model of an air-space-ground integrated Internet of Things data acquisition system; the wireless communication model for constructing an air-space-ground integrated Internet of Things data acquisition system includes K LEO satellites, N multi-rotor UAV base stations and M IoT devices, a ground-to-air access link is established between the IoT device and the UAV base station, and an air-to-space return link is established between the UAV base station and the LEO satellite; a second module, for executing S2, characterizing the throughput of the ground-to-air access link and the air-to-space return link during data acquisition; a third module, for executing S3, constructing a resource allocation model in an air-space-ground integrated Internet of Things data acquisition system with an optimization problem P1 that maximizes the minimum throughput of all ground-to-air access links; a fourth module, for executing S4, introducing non-negative auxiliary variables, equivalently converting the optimization problem P1 into the optimization problem P2; decomposing the optimization problem P2 into sub-problems, and solving the resource allocation model in the air-space-ground integrated Internet of Things data acquisition system according to a preset efficient algorithm. As used above, the term "module" can implement a combination of software and / or hardware that implements a predetermined function. Although the system described in the above embodiment is preferably implemented in software, implementation in hardware, or a combination of software and hardware is also possible and conceived. For parts that are not described in detail in each module, please refer to the relevant description of this embodiment.

[0067] According to a third aspect of an embodiment of the present invention, a processor is provided, which is used to run a program. When the program is running, the modules of the resource allocation method in the integrated air-space-ground Internet of Things data acquisition system described in any one of the above are executed, and each module is located in the same processor; or, the above modules are located in different processors in any combination.

[0068] Embodiment 2: The following is an optional specific implementation of the present invention in conjunction with the accompanying drawings:

[0069] like Figure 1-Figure 4 As shown, a resource allocation method in an air-ground-integrated Internet of Things data acquisition system includes:

[0070] S1. Build a wireless communication model for the integrated air-ground-space Internet of Things data acquisition system and initialize system parameters;

[0071] The wireless communication model for constructing the air-ground integrated IoT data acquisition system includes K LEO satellites located in a specific orbit, N dynamically movable multi-rotor UAV base stations, and M IoT devices fixed on the ground. A ground-to-air access link is established between the IoT device and the UAV base station, and an air-to-space return link is established between the UAV base station and the LEO satellite.

[0072] Specifically: For the wireless communication model of the integrated air-ground-space IoT data acquisition system, both the ground-to-air access link and the air-to-space backhaul link use frequency division multiple access technology for wireless communication. The LEO satellite, UAV base station and IoT device set are defined as and Among them, the number of M is much larger than N. In the three-dimensional Cartesian space coordinate system, the spatial positions of the kth LEO satellite, the nth UAV base station and the mth IoT device are defined as and

[0073] Exemplarily, the initialization system parameters include the initial position coordinates of the IoT device, the initial position coordinates of the UAV base station, the position coordinates of the LEO satellite, the maximum flight speed and maximum energy consumption of the UAV base station, the maximum and minimum flight altitudes of the UAV base station, the horizontal flight radius of the UAV base station, the minimum safety distance between any two UAV base stations, etc.

[0074] S2, characterizes the throughput of the ground-to-air access link and the air-to-space backhaul link during data collection;

[0075] The throughput of the ground-to-air access link during the characterization data collection process includes:

[0076] According to the positional relationship between the UAV base station and the IoT device, the generalized probability propagation model is used to calculate the channel gain between the nth UAV base station and the mth IoT device based on the consideration of both the line-of-sight propagation channel and the non-line-of-sight propagation channel. in, and are the probabilities of line-of-sight propagation channel and non-line-of-sight propagation channel respectively, e is a natural constant, ζ 1 and 2 are system parameters related to the propagation environment, is the elevation angle from the mth IoT device to the nth UAV base station, is the height of the nth UAV base station, is the path cost of the line-of-sight propagation channel, is the path consumption of the non-line-of-sight propagation channel, ε is the path consumption factor in the free space propagation environment, η L and η NLare the additional path consumption factors under line-of-sight propagation channel and non-line-of-sight propagation channel respectively; the proportion of 6GHz band bandwidth allocated to the mth IoT device is The transmission power of the mth IoT device is given as represents the time when the nth UAV base station collects IoT device data and sends it to the LEO satellite. The variance of the additive white Gaussian noise on each UAV base station side is Under a given power condition, combined with Shannon's formula, the throughput of the ground-to-air access link is It is expressed as:

[0077]

[0078] Among them, B G2A Indicates the bandwidth of the ground-to-air access link.

[0079] The throughput of the space-to-air backhaul link during the characterization data collection process includes:

[0080] According to the positional relationship between the LEO satellite and the UAV base station, combined with the Rician fast fading model, the channel gain between the nth UAV base station and the kth LEO satellite is obtained. Where c is the speed of light, f c is the carrier frequency, is the fast fading part of the channel gain between the nth UAV base station and the kth LEO satellite, is the distance between the nth UAV base station and the kth LEO satellite; the ratio of the millimeter wave frequency band bandwidth allocated to the nth UAV base station is The transmission power of the given nth UAV base station is represents the time when the nth UAV base station collects IoT device data and sends it to the LEO satellite; the variance of the additive white Gaussian noise on each LEO satellite side is Under a given power condition, combined with Shannon's formula, the throughput of the space-to-space backhaul link is It is expressed as:

[0081]

[0082] Among them, B A2S Indicates the bandwidth of the space backhaul link.

[0083] S3, construct a resource allocation model in the air-ground-integrated IoT data collection system;

[0084] The resource allocation model in the air-ground-integrated Internet of Things data acquisition system is specifically constructed as follows: in the air-ground-integrated Internet of Things data acquisition system, by jointly optimizing the multi-dimensional communication resources of three-dimensional node pairing, bandwidth allocation and UAV base station layout, the resource allocation model in the air-ground-integrated Internet of Things data acquisition system is constructed for the optimization problem P1 to maximize the minimum throughput of all ground-air access links; in the above, the three-dimensional node pairing variable is The bandwidth allocation variable is UAV base station layout variables involve the spatial location of the UAV base stations Flight speed Data collection time

[0085] The resource allocation model in the air-ground-integrated IoT data collection system is expressed as:

[0086]

[0087] Where, the goal of the optimization model P1 is to maximize the minimum throughput of all ground-to-air access links; constraint C1 represents the bandwidth allocation variable of IoT devices and UAV base stations is between 0 and 1; Constraint C2 requires that the sum of IoT device bandwidth allocation variables and the sum of UAV base station bandwidth allocation variables cannot exceed 1; Constraint C3 requires that the IoT device attachment optimization variable used for 3D node pairing Drone attachment optimization variables All are binary values ​​(0 for unpaired, 1 for paired); Constraint C4 requires that each IoT device can only connect to one UAV base station at most, and each UAV base station can only connect to one LEO satellite; Constraint C5 requires that the throughput of the ground-to-air access link for any UAV base station is less than or equal to the throughput of the air-to-space return link; Constraint C6 requires that the flight speed of the UAV base station Less than or equal to Constraints C7-C9 require that the horizontal coordinates of each UAV base station are within a circular area with a radius of r, and the flight altitude has an upper limit z max and the lower limit z min , and the distance between any two UAV base stations is not less than the safety distance d min , Constraint C10 requires that the energy consumption of each UAV base station be capped at Q max , represents the initial position of all UAVs; the energy consumption of the UAV base station is mainly determined by the propulsion part, and the expression for the relationship between the flight speed and the energy consumption of the UAV base station is constructed as follows:

[0088]

[0089] Where P0 and P I They represent the UAV base station rotor blade profile power and induced power, U tip is the tip speed of the rotor blade, v 0 represents the average induced velocity of the UAV base station when hovering, D 0 ,ρ,S,A represent the fuselage drag coefficient, air density, rotor solidity, and rotor disc area, respectively. It should be noted that Q(0) represents the energy consumption of each UAV to collect data when it is stationary in the hovering position.

[0090] S4. Design an efficient algorithm to solve the resource allocation model in the integrated air-space-ground Internet of Things data acquisition system.

[0091] For the original mixed integer non-convex optimization problem P1, a non-negative auxiliary variable θ is introduced to convert the optimization problem P1 into the optimization problem P2:

[0092]

[0093] Wherein, constraint C11 requires that the throughput of any ground-to-air access link is greater than or equal to a non-negative auxiliary variable.

[0094] It is easy to know that P2 is still a mixed integer non-convex optimization problem. In order to obtain an effective approximate solution to the problem in polynomial time, the present invention proposes a low-complexity and efficient mechanism. Through the alternating optimization theory, the optimization problem P2 is decomposed into three-dimensional node pairing sub-problems, bandwidth allocation sub-problems, and UAV base station layout sub-problems for alternating optimization.

[0095] S4.1. When the bandwidth allocation variables and the UAV base station layout are given, the remaining three-dimensional node pairing subproblem can be constructed as the optimization problem P3:

[0096]

[0097] In the formula, the objective function and constraints are linear, the variables are binary integer variables, and the optimization problem P3 is a linear integer programming problem. With the help of mature linear integer programming tools, the global optimal solution can be obtained in multiple times. The optimal solution is Exemplarily, the linear integer programming tool may be CPLEX, Gurobi, etc.

[0098] S4.2. When the three-dimensional node pairing and UAV base station layout are given, the remaining bandwidth allocation subproblem can be formulated as the optimization problem P4:

[0099]

[0100] Wherein, constraint C5 is an inequality that two concave functions are less than or equal to 0, so the feasible solution set of the variables does not constitute a convex set, so problem P4 is a non-convex optimization problem.

[0101] In view of the above characteristics, the present invention adopts a continuous convex approximation algorithm to solve the optimization problem P4, and the solution steps are as follows:

[0102] S4.21, set the current iteration number i=1, the maximum iteration number is 1, and the error threshold is ε 1 , the bandwidth allocation ratio of IoT devices that meets the constraints is initialized as

[0103] S4.22, in At the point, the constraint condition C5 in the optimization problem P4 is expanded by a first-order Taylor, and an approximate convex optimization problem P5 is constructed. The current optimal solution is obtained by using the interior point method.

[0104] S4.23, if i ≥ 2 and the performance difference of adjacent iterations is less than the first error threshold, that is, θ (i) -θ (i-1) ≤ε 1 , then the algorithm converges, and the current is the optimal solution; otherwise, update the Lagrangian expansion point Set i=i+1 and return to S4.22; in addition, if the number of iterations exceeds the maximum number 1, stop the iteration and return the current solution.

[0105] Specifically, for an iteration process, the specific description is as follows:

[0106] definition And Taking the second-order partial derivative we get:

[0107]

[0108] This proves About is a concave function, so The global upper bound of Right Perform a first-order Taylor expansion, that is:

[0109]

[0110] In the formula, express exist The first-order derivative at , Indicates the bandwidth allocation ratio of the mth IoT device after the i-th iteration.

[0111] Therefore, the optimization problem P4 can be further written as the optimization problem P5:

[0112]

[0113] Constraints for Optimization Problem P5 The feasible domain of and C9 is a convex set, and the objective function and other constraints are linear. Therefore, problem P5 is a convex optimization problem, and the present invention adopts the interior point method to solve it.

[0114] Since the optimization problem P5 is an approximate transformation of the optimization problem P4, it still needs to be optimized alternately. Therefore, the present invention gradually approaches the optimal solution of the non-convex optimization problem P4 by continuously optimizing the approximate convex optimization problem P5, and obtains a suboptimal solution to the optimization problem P4.

[0115] S4.3. When the three-dimensional node pairing and bandwidth allocation variables are given, the remaining UAV base station layout subproblems are constructed as optimization problem P6:

[0116]

[0117] It is easy to see that the UAV base station layout subproblem P6 is still a non-convex optimization problem, and the present invention uses sequential quadratic programming to solve it. For the convenience of explanation, further, based on the non-negative auxiliary variable θ, the optimization problem P6 is rewritten as the optimization problem P7:

[0118]

[0119] In the formula, represents the opposite of the objective function in the optimization problem P6; Represents the vector of all variables to be optimized in problem P6; represents the total number of constraints in problem P6; It represents the simplified form of the constraints of problem P6. The Lagrangian function of the optimization problem P7 is expressed as:

[0120]

[0121] In the formula, λ=[λ 1 ,λ 2 ,...,λ I ] Τ is the Lagrange multiplier, λ i represents the non-negative Lagrange multiplier of the ith constraint.

[0122] The core idea of ​​sequential quadratic programming is: for the original nonlinear optimization problem, through iterative optimization, it is transformed into a quadratic programming problem for approximate solution at each iteration, and the optimal solution of the current quadratic programming problem is used as a new search point to construct a new quadratic programming sub-problem. The specific steps are as follows:

[0123] At a given initial search point At , the optimization problem P7 is converted into an approximate quadratic programming subproblem P8:

[0124]

[0125] In the formula, and Respectively represent the objective function at the current iteration point The gradient at; T represents transpose; H (j) Indicates the Lagrangian function at the current iteration point The approximate Hessian matrix at ; Indicates that the i-th constraint condition is at the current iteration point The value at is the set of slack variables for problem P8, ensuring the feasibility of the constraints.

[0126] Finally, the constructed approximate quadratic programming subproblem P8 is solved by the interior point method, and based on the current optimal solution Use line search to update the next iteration search point (i.e. the search point of the j+1th iteration), construct the approximate quadratic programming subproblem again, and obtain the optimization result of the drone variable subproblem by iterative optimization until convergence is met.

[0127] In order to ensure the solvability of the optimization problem P8 and the global convergence of the algorithm, the approximate Hessian matrix H is first updated by the damped Broyden-Fletcher-Goldfarb-Shanno (BFGS) method. (j+1) Then, the step size is adjusted by combining Fletcher's augmented Lagrange value function and the optimal solution of the current problem P8 to determine the next search point The detailed steps are as follows:

[0128] Step 1: Update the approximate Hessian matrix H (j) ; First, use the interior point method to search for the point in the current iteration The global optimal solution of problem P8 is obtained The corresponding Lagrange dual variable is Then further define:

[0129]

[0130] r (j) =φ (j) y (j) +(1-φ (j) )H (j) ω (j)

[0131] In the formula, the next iteration search point based on and get; represents the Lagrangian function with respect to the variable The first derivative of (j+1) Depend on Or the least squares estimate is updated; φ (j) represents the damping factor, which is defined as:

[0132]

[0133] Based on the above definition, the approximate positive definite Hessian matrix H is obtained by the damped BFGS formula (j+1) :

[0134]

[0135] Step 2: Update the search point for the next iteration Combine the Fletcher augmented Lagrange value function and the optimal solution of the current problem P8 to adjust the step size. Fletcher augmented Lagrange value function Defined as:

[0136]

[0137] In the formula, u=[u 1 ,u 2 ,...,u I ] Τ represents the set of slack variables for problem P7, and ν>0 represents an adjustable penalty coefficient. Through practice, it is found that when ν is small enough, the optimal solution to the current problem P8 is The Fletcher augmented Lagrange value function at the current iteration point The descending direction of .

[0138] In order to ensure the update step size and avoid oscillation, the line search method is used to update the step size a (j) . First, initialize the step size a (j) = 1, and through the factor κ a ∈(0,κ) adjust the step size a (j) (a (j) =κ a a (j) ), until the Fletcher augmented Lagrange value function is satisfied Achieve adequate descent and satisfy:

[0139]

[0140] In the formula, Indicates that at the current step length a (j) Fletcher augmented Lagrange cost function at , η is the model parameter, It means that the value function is The first-order derivative at , when the line search method converges, the update method of the next iterative search point is given by the following formula:

[0141]

[0142] Based on the above description, the specific implementation of the steps described in the entire S4 can be summarized in S4.4 as follows:

[0143] Initialization: Set t=1 to the current number of iterations, T max Set to the maximum number of iterations. Given the initial variables that satisfy the constraints of problem P2

[0144] Step 1: In the current optimization variable Under these conditions, the suboptimal solution of the three-dimensional node pairing subproblem is obtained through S4.1 optimization and updated

[0145] Step 2: In the current optimization variable Under these conditions, the suboptimal solution of the current bandwidth allocation subproblem is obtained through S4.2 optimization and updated

[0146] Step 3: In the current optimization variable Under these conditions, the suboptimal solution of the current UAV base station layout subproblem is obtained through S4.3 optimization and updated

[0147] Step 4: If t ≥ 2 and θ (t) -θ (t-1) ≤ε 2 , then the proposed algorithm converges, and the current This is the suboptimal solution to problem P1; otherwise, set t=t+1 and return to Step 1.

[0148] Further, combined with experimental data, the following is explained:

[0149] The present invention considers that in a circular area with a radius of 200m, IoT devices are randomly distributed in the circular area, UAV base stations are evenly distributed directly above the circular area (with a height between 20m and 100m), three near-earth satellites are directly above the circular area, and the horizontal projection is offset 15° to the left and right relative to the center of the area (with a height of 7×10 5m). The number of IoT devices M = 30, the number of UAV base stations N = 4. Other system parameter settings are shown in Table 1.

[0150] Table 1 Parameter setting table

[0151]

[0152]

[0153] Figure 3 The convergence analysis of the algorithm of the present invention under 100 random network topologies is demonstrated. Taking maximizing the minimum throughput of all data acquisition links as the objective function, it can be observed that with the increase in the number of iterations, the objective function value (utility value) gradually increases and tends to be stable, indicating that the algorithm can converge under the conditions of different numbers of IoT devices. However, as the number of IoT devices increases, the system communication resources are constrained, resulting in a gradual decrease in the final convergence value of the objective function. This reflects the restrictive impact of resource allocation on performance when the number of devices increases, and also verifies the applicability of the algorithm in resource-constrained scenarios.

[0154] Figure 4 The present invention is compared with the following two comparison mechanisms under 100 random network topologies:

[0155] Comparison mechanism 1: For the binary integer auxiliary variables of IoT devices and UAV base stations, they are relaxed to continuous variables between (0,1), and the converted continuous non-convex optimization problem P1 is solved using the sequential quadratic programming method to obtain a suboptimal solution.

[0156] Comparison mechanism 2: Evenly distribute the bandwidth share of each IoT device and UAV base station, and the rest remains consistent with the algorithm proposed in the present invention.

[0157] Depend on Figure 4 It can be seen that compared with the comparative mechanism 1, the proposed mechanism (algorithm 1) of the present invention decomposes the problem into three sub-parts and designs an efficient algorithm for each sub-problem to solve it. This strategic decomposition significantly improves the system performance, with an average performance improvement of 26.08bps / Hz, which fully proves the effectiveness of the proposed method in improving the overall system performance; compared with the comparative mechanism 2, the proposed algorithm achieves a significant average performance improvement of 95.61bps / Hz. This significant improvement shows that the proposed algorithm has obvious advantages in bandwidth allocation optimization, especially in comparison with the traditional uniform bandwidth allocation strategy, showing higher system performance.

[0158] The specific implementation modes of the present invention are described in detail above in conjunction with the accompanying drawings, but the present invention is not limited to the above implementation modes, and various changes can be made within the knowledge scope of ordinary technicians in this field without departing from the purpose of the present invention.

Claims

1. A resource allocation method in an air-ground-integrated Internet of Things data acquisition system, characterized in that: The following steps are involved: S1. Construct a wireless communication model of an air-ground-integrated IoT data collection system; the wireless communication model for constructing an air-ground-integrated IoT data collection system includes K LEO satellites, N multi-rotor UAV base stations, and M IoT devices, a ground-to-air access link is established between the IoT device and the UAV base station, and an air-to-space return link is established between the UAV base station and the LEO satellite; S2, characterizes the throughput of the ground-to-air access link and the air-to-space backhaul link during data collection; S3, constructing a resource allocation model in the air-ground integrated IoT data collection system based on the optimization problem P1 of maximizing the minimum throughput of all ground-air access links; S4. Introduce non-negative auxiliary variables to convert the optimization problem P1 into the optimization problem P2; decompose the optimization problem P2 into sub-problems, and solve the resource allocation model in the integrated air-space-ground Internet of Things data acquisition system based on the preset efficient algorithm.

2. The resource allocation method in the air-ground-integrated Internet of Things data acquisition system according to claim 1 is characterized in that: The throughput of the ground-to-air access link and the air-to-space backhaul link during the characterization data collection process includes: According to the positional relationship between UAV base stations and IoT devices, a generalized probability propagation model is used to calculate the channel gain between each UAV base station and IoT device while considering both line-of-sight and non-line-of-sight propagation channels. Based on the channel gain between each UAV base station and IoT device, the throughput of the ground-to-air access link during data collection is calculated in combination with the Shannon formula. According to the positional relationship between LEO satellites and UAV base stations, combined with the Rician fast fading model, the channel gain between each LEO satellite and the UAV base station is calculated; according to the channel gain between each LEO satellite and the UAV base station, combined with the Shannon formula, the throughput of the space-to-space backhaul link during the data acquisition process is calculated.

3. The resource allocation method in the air-ground-integrated Internet of Things data acquisition system according to claim 1 is characterized in that: The resource allocation model in the integrated air-space-ground-integrated Internet of Things data acquisition system is specifically constructed as follows: in the integrated air-space-ground-integrated Internet of Things data acquisition system, by jointly optimizing the multi-dimensional communication resources of three-dimensional node pairing, bandwidth allocation and UAV base station layout, an objective function and constraint conditions are established for the optimization problem P1 to maximize the minimum throughput of all ground-to-air access links to construct a resource allocation model in the integrated air-space-ground-integrated Internet of Things data acquisition system.

4. The resource allocation method in the air-ground-integrated Internet of Things data acquisition system according to claim 3 is characterized in that: The resource allocation model in the air-ground-integrated IoT data collection system is expressed as: Where, the goal of the optimization problem P1 is to maximize the minimum throughput of all ground-to-air access links; constraint C1 represents the bandwidth allocation variable of IoT devices and UAV base stations is between 0 and 1; Constraint C2 requires that the sum of IoT device bandwidth allocation variables and the sum of UAV base station bandwidth allocation variables cannot exceed 1; Constraint C3 requires that the IoT device attachment optimization variable used for 3D node pairing Drone attachment optimization variables All are binary values; Constraint C4 requires that each IoT device can only connect to one UAV base station at most, and each UAV base station can only connect to one LEO satellite; Constraint C5 requires that the throughput of the ground-to-air access link for any UAV base station is less than or equal to the throughput of the air-to-space backhaul link; Constraint C6 requires the UAV base station to fly at a speed of Less than or equal to ζ max ; Constraints C7-C9 require that the horizontal coordinates of each UAV base station are within a circular area with a radius of r, and the flight altitude has an upper limit z max and the lower limit z min , and the distance between any two UAV base stations is not less than the safety distance d min , Constraint C10 requires that the energy consumption of each UAV base station be capped at Q max , Indicates the initial position of all UAV base stations; the spatial position of the nth UAV base station Q(0) represents the energy consumption of each UAV in collecting data when it is stationary in the hovering position; Indicates the flight speed of the UAV base station is The energy consumption of collecting data.

5. The resource allocation method in the air-ground-integrated Internet of Things data acquisition system according to claim 1, characterized in that: The optimization problem P2 is decomposed into sub-problems, specifically including: a three-dimensional node pairing sub-problem, a bandwidth allocation sub-problem, and a UAV base station layout sub-problem.

6. The resource allocation method in the air-ground-integrated Internet of Things data acquisition system according to claim 5, characterized in that: The method of solving the resource allocation model in the air-ground-integrated Internet of Things data acquisition system according to a preset efficient algorithm includes: S4.

1. When the bandwidth allocation and UAV base station layout variables are given, the remaining three-dimensional node pairing subproblems can be constructed as optimization problem P3, which is solved by linear integer programming tools; S4.2, when the three-dimensional node pairing and UAV base station layout are given, the remaining bandwidth allocation sub-problem can be constructed as optimization problem P4, and the continuous convex approximation algorithm is used to write the optimization problem P4 as optimization problem P5 for optimization solution; S4.3, when the three-dimensional node pairing and bandwidth allocation variables are given, the remaining UAV base station layout subproblems are constructed as optimization problem P6, which is solved using the sequential quadratic programming algorithm; S4.

4. Combine S4.1-S4.3 above and use the alternating optimization method to obtain the resource allocation result of the integrated air-space-ground-integrated Internet of Things data acquisition system.

7. The resource allocation method in the air-ground-integrated Internet of Things data acquisition system according to claim 6, characterized in that: The S4.4 is specifically: Initialization: Set t=1 to the current number of iterations, T max Set to the maximum number of iterations; given the initial variables that satisfy the constraints of the optimization problem P2 After initialization, perform the following steps: Step 1: In the current optimization variable Under these conditions, the suboptimal solution of the three-dimensional node pairing subproblem is obtained through S4.1 optimization and updated to Step 2: In the current optimization variable Under these conditions, the suboptimal solution of the current bandwidth allocation subproblem is obtained through S4.2 optimization and updated to Step 3: In the current optimization variable Under these conditions, the suboptimal solution to the UAV base station layout subproblem is obtained through S4.3 optimization and updated to Step 4: If t ≥ 2 and θ (t) -θ (t-1) ≤ε2, the proposed algorithm converges, and the current This is the suboptimal solution to the optimization problem P1; otherwise, set t=t+1 and return to Step 1.

8. A resource allocation system in an air-ground-integrated Internet of Things data acquisition system, characterized in that: A module for the resource allocation method in the air-space-ground integrated Internet of Things data acquisition system as described in any one of claims 1-7.

9. A processor, the processor being used to run a program, characterized in that: When the program is running, the resource allocation method in the air-space-ground integrated Internet of Things data acquisition system described in any one of claims 1 to 7 is executed.

Images