Drone Deployment Method and System Based on Dense Boundary Priority Service

Abstract

The present invention belongs to the technical field of mobile edge computing, and specifically relates to a method and system for deploying unmanned aerial vehicles (UAVs) based on dense boundary priority service. The method includes: S1, obtaining relevant information in the current scenario; S2, establishing a system optimization model with the UAV positions and the communicable channel gains as constraints and maximizing the average transmission rate as the objective; S3, decomposing the optimization problem into a maximum service radius and vertical position sub-problem and a region division and horizontal position sub-problem; S4, solving the maximum service radius and vertical position sub-problem; S5, solving the UAV horizontal position sub-problem. The present invention has the characteristics that in a complex urban environment, the UAV can be used as a mobile edge server to assist users or edge servers in processing tasks when the data volume surges, so as to maximize the average transmission rate of the system.

Description

Technical Field

[0001] The present invention belongs to the technical field of mobile edge computing, and particularly relates to a method and system for deploying unmanned aerial vehicles (UAVs) based on dense boundary priority service. Background Art

[0002] Mobile edge computing alleviates the pressure on user equipment by providing highly reliable and low-latency services to users, and is envisioned as one of the key technologies for the next-generation mobile network. Research on MEC, such as computing offloading, resource allocation, network deployment, etc., has received increasing attention. Traditional MEC cannot be flexibly deployed due to its fixed architecture. With the exponential growth of user data, the deployment cost of mobile edge servers has become an issue worthy of consideration. Moreover, it is difficult to deploy in remote areas such as deserts and forests. Therefore, traditional MEC is difficult to cope with scenarios where user locations and service requirements are constantly changing.

[0003] In recent years, due to its high mobility, flexible deployment, low cost, etc., unmanned aerial vehicles (UAVs) have made up for the disadvantages of mobile edge servers in deployment. More and more research uses UAVs as aerial base stations to assist traditional MEC in collaborative work. Since UAVs have high mobility, they can work in collaboration with nearby mobile edge servers, greatly reducing the additional communication latency consumed by user equipment due to long-distance transmission. In addition, UAVs can also act as mobile relays to further improve network throughput and coverage. When traditional ground base stations are damaged, UAVs can act as temporary aerial base stations to establish emergency communications. For this, how to deploy multiple UAVs in a complex urban environment to provide users with high-quality service quality has become an urgent problem to be solved.

[0004] Based on the above problems, it is very important to design a method and system for deploying UAVs based on dense boundary priority service for a complex urban environment, which can use UAVs as mobile edge servers to assist users or edge servers in processing tasks when the data volume surges, so as to maximize the average transmission rate of the system.

[0005] For example, the Chinese patent document with the application number CN202010612159.4 describes a method and terminal device for deploying relay UAVs. The method includes: taking the minimum number of relay UAVs as the objective function, and taking the effective communication distance constraint between communication nodes and the safety distance constraint between communication nodes as the constraint conditions to establish a relay UAV deployment model; solving the relay UAV deployment model to obtain the deployment positions of each relay UAV. Although it can effectively complete the deployment of relay UAVs with the minimum number of relay UAVs and has strong practicability, the disadvantage is that the above method cannot be applied to a complex urban environment and cannot achieve the technical effect of maximizing the average transmission rate of the system. Summary of the Invention

[0006] The present invention aims to overcome the problems in the prior art that the existing UAV deployment methods cannot be applied to complex urban environments and cannot provide users with high-quality service quality. A UAV deployment method and system based on dense boundary priority service are provided for complex urban environments, which can use UAVs as mobile edge servers to assist users or edge servers in processing tasks when the data volume surges, maximizing the average transmission rate of the system.

[0007] To achieve the above object of the invention, the following technical solutions are adopted:

[0008] The UAV deployment method based on dense boundary priority service includes the following steps:

[0009] S1, initialization phase:

[0010] Obtain relevant information in the current scenario, including ground user location information, the corresponding urban model information in the current environment, and the minimum channel gain that can communicate in the current environment;

[0011] S2, establish an optimization model:

[0012] According to the overall optimization goal, establish a system optimization model with the UAV position and the communicable channel gain as constraints and the maximization of the average transmission rate as the goal;

[0013] S3, decompose the optimization problem into a maximum service radius and vertical position sub-problem and a region division and horizontal position sub-problem;

[0014] S4, solve the maximum service radius and vertical position sub-problem based on the KKT condition;

[0015] S5, solve the UAV horizontal position sub-problem based on the region division method of dense boundary priority service.

[0016] Preferably, step S1 includes the following steps:

[0017] S11, define the UAV set as A = {A1, A2,..., A m ,..., A M}, and the user set as U = {u1, u2,..., u k ,..., u K}; assume that all UAVs have the same vertical height, and use a three-dimensional Cartesian coordinate system to represent the coordinates of the mth UAV as (x m , y m , h), and the coordinates of the kth user as (a k , b k,0), assume that all user positions are fixed; the following restrictions are imposed on the UAV position:

[0018] X min ≤x m ≤X max (44)

[0019] Y min ≤y m ≤Y max (45)

[0020] H min ≤h≤H max (46)

[0021] where X min 、X max 、Y min 、Y max 、H min 、H max represent the minimum and maximum ranges of the horizontal and vertical positions of the UAV, respectively;

[0022] S12, establish a LoS channel model in the urban environment. Under the LoS channel, the communication probability between UAV A m and mobile user u k is:

[0023]

[0024] where a and b are constant values depending on the current environment, represents the angle between UAV A m and mobile user u k ;

[0025] Under the LoS channel, the path loss between UAV A m and mobile user u k is expressed as:

[0026]

[0027] where, represents the distance between UAV A m and mobile user u k , f c represents the carrier frequency, c represents the speed of light, and η LoS represents the additional path loss corresponding to the LoS channel, which is a constant value depending on the environment;

[0028] S13, under the NLoS channel, the communication probability between UAV A m and mobile user u k is:

[0029]

[0030] In the NLoS channel, the communication path loss is expressed as:

[0031]

[0032] where η NLoS represents the additional path loss corresponding to the NLoS channel and is a constant value depending on the environment;

[0033] It is obtained that the path loss between UAV A m and the mobile user u k is expressed as:

[0034]

[0035] It is obtained that the path loss between UAV A m and the mobile user u k The transmission rate is expressed as

[0036]

[0037] where B represents the system transmission bandwidth, P represents the average transmit power of the user, and σ 2 represents Gaussian white noise;

[0038] Using the minimum channel gain g0 to represent the condition for the channel to be communicable, when the condition

[0039]

[0040] is satisfied, communication can be established between UAV A m and the mobile user u k

[0041] Preferably, step S2 includes the following steps:

[0042] S21, the system optimization model is modeled as follows:

[0043]

[0044] s.t. X min ≤x m ≤X max (55)

[0045] Y min ≤y m ≤Y max (56)

[0046] H min ≤h≤H max (57) ​

[0047]

[0048] Preferably, step S4 includes the following steps:

[0049] S41. Derive from the minimum channel gain that can be communicated between the UAV and the mobile user:

[0050] PL m,k ≤ -10log(g0) (59)

[0051] Furthermore, obtain the maximum service radius r of the UAV and the corresponding height h;

[0052] The sub - problem representation of the UAV's vertical position and the maximum service radius is obtained as:

[0053]

[0054] s.t. H min ≤ h ≤ H max (61)

[0055]

[0056] S42. Use the trigonometric relationship h = rtanθ m,n , and convert the maximum service radius r of the UAV and the corresponding height h into the relationship between the maximum service radius r of the UAV and the elevation angle θ m,n between them, and convert the problem into:

[0057]

[0058] s.t. PL m,k +10log(g0) ≤ 0 (64)

[0059] H min - rtanθ m,k ≤ 0 (65)

[0060] rtanθ m,k - H max ≤ 0 (66)

[0061] According to the above - mentioned optimization problem, it is expressed by the Lagrangian function as:

[0062]

[0063] where λ1, λ2, λ3 are Lagrange multipliers;

[0064] S43. According to the KKT conditions, the parameters satisfy:

[0065]

[0066] λ2(H min -rtanθ m,k )=0 (70)

[0067] λ3(rtanθ m,k -H max )=0 (71)

[0068]

[0069] H min -rtanθ m,k ≤0 (74)

[0070] rtanθ m,k -H max ≤0 (75)

[0071] λ1≥0, λ2≥0, λ3≥0 (76)

[0072] In summary, it is divided into three cases to derive the relevant expressions of r and θ m,k as follows:

[0073]

[0074] where θ m,k represents the angle between the drone A m and the mobile user u k and

[0075] For the above three cases, under the condition of satisfying the λ2 and λ3 constraint conditions, the corresponding r and θ are solved m,k and compared, and the solution with the largest service radius r is the final solution.

[0076] Preferably, step S5 includes the following steps:

[0077] S51, through the solution of step S4, obtain the maximum service radius and the corresponding height of the drone in the current environment. At this time, the optimization model is simplified to:

[0078]

[0079] s.t. X min ≤x m ≤X max (79)

[0080] Y min ≤y m ≤Y max (80)

[0081]

[0082] For a given area, solve the problem of deploying multiple drones. By using an algorithm based on dense boundary priority service, divide the overall area into multiple partial areas, and solve the problem of deploying a single drone within the divided areas; for the partial area, the expression in (35) is:

[0083]

[0084]

[0085] According to equation (9), R m,k increases as increases. For equation (39), it is equivalent to finding

[0086]

[0087] For equation (41), in a general urban environment, η LoS < η NLoS . It is concluded that when h is determined, decreases as L = (x m - a k ) 2 + (y m - b k ) 2 increases. It is concluded that solving equation (39) is transformed into solving:

[0088]

[0089] That is, solve for the position within the given area that minimizes the sum of the distances from the drone to all users, which is the position of drone deployment.

[0090] By solving the Hessian matrix corresponding to equation (42):

[0091]

[0092] It is obtained that a1 = 2, a1a4 - a2a3 = 4, and the solution is obtained through the CVX optimization tool.

[0093] The present invention also provides a drone deployment system based on dense boundary priority service, including:

[0094] An initialization module for obtaining relevant information in the current scenario, including ground user position information, the corresponding urban model information in the current environment, and the minimum channel gain that can communicate in the current environment;

[0095] An optimization model building module, which is used to build a system optimization model with the position of the unmanned aerial vehicle (UAV) and the gain of the communicable channel as constraints and the maximization of the average transmission rate as the objective according to the overall optimization objective;

[0096] An optimization problem decomposition module, which is used to decompose the optimization problem into a maximum service radius and vertical position sub-problem and a region division and horizontal position sub-problem;

[0097] A vertical position solving module, which is used to solve the maximum service radius and vertical position sub-problem based on the KKT conditions;

[0098] A horizontal position solving module, which is used to solve the UAV horizontal position sub-problem based on the region division method of serving the dense boundary first.

[0099] Compared with the prior art, the beneficial effects of the present invention are as follows: (1) The present invention optimizes the three-dimensional positions of multiple UAVs in a UAV-assisted mobile edge network, and realizes the objective of maximizing the system average transmission rate under the condition of limited UAV resources; (2) The UAV deployment method implemented by the present invention only depends on the current urban model and the user distribution, and this parameter can be obtained in the initialization stage. Moreover, compared with the traditional heuristic method, the method proposed by the present invention does not depend on the selection of the initial solution, and can obtain the UAV deployment positions that can make the system average transmission rate relatively large. Description of the Drawings

[0100] Figure 1 A network example diagram of the UAV deployment method based on serving the dense boundary first provided by an embodiment of the present invention;

[0101] Figure 2 A flowchart of the UAV deployment method based on serving the dense boundary first provided by an embodiment of the present invention;

[0102] Figure 3 A UAV position distribution diagram under the UAV deployment method based on serving the dense boundary first provided by an embodiment of the present invention;

[0103] Figure 4 A quantity diagram of the required UAVs under the dense boundary first service method and the K-means method provided by an embodiment of the present invention;

[0104] Figure 5 A comparison diagram of the system average transmission rates under the dense boundary first service method and the K-means method provided by an embodiment of the present invention. Detailed Embodiments

[0105] To more clearly illustrate the embodiments of the present invention, the following will describe the specific implementation manners of the present invention with reference to the accompanying drawings. Obviously, the accompanying drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings, and other implementation manners can also be obtained.

[0106] Embodiment:

[0107] As Figure 2 shown, the method for deploying drones based on dense boundary priority service of the present invention includes the following steps:

[0108] S1, initialization phase:

[0109] Obtain relevant information in the current scenario, including ground user location information, the corresponding urban model information in the current environment, and the minimum channel gain that can communicate in the current environment;

[0110] S2, establish an optimization model:

[0111] According to the optimized overall objective, establish a system optimization model with the drone position and the communicable channel gain as constraints and the maximization of the average transmission rate as the objective;

[0112] S3, decompose the optimization problem into a maximum service radius and vertical position sub-problem and a region division and horizontal position sub-problem;

[0113] S4, solve the maximum service radius and vertical position sub-problem based on the KKT conditions;

[0114] S5, solve the drone horizontal position sub-problem based on the region division method of dense boundary priority service.

[0115] The specific process of the present invention is as follows:

[0116] The system model of the present invention is as Figure 1 shown. Consider a wireless communication network composed of M drone nodes and K users. Each drone can serve multiple users. Each user must be under the coverage of a drone node and can establish communication with the drone. Define the drone set as A = {A1, A2,..., A m ,..., A M}, the user set as U = {u1, u2,..., u k ,..., u K}. Assume that all drones have the same vertical height. Using a three-dimensional Cartesian coordinate system, the coordinates of the mth drone can be expressed as (x m , y m , h), and the coordinates of the kth user can be expressed as (a k , bk , 0), this chapter assumes that all user positions are fixed. There are the following restrictions on the UAV position:

[0117] X min ≤ x m ≤ X max (87)

[0118] Y min ≤ y m ≤ Y max (88)

[0119] H min ≤ h ≤ H max (89)

[0120] Among them, X min , X max , Y min , Y max , H min , H max respectively represent the minimum and maximum ranges of the horizontal and vertical positions of the UAV.

[0121] This invention considers the influence of the environment on the LoS channel, establishes a LoS channel model in an urban environment. Under the LoS channel, the communication probability between UAV A m and mobile user u k is

[0122]

[0123] where a and b are constant values depending on the current environment, represents the included angle between UAV A m and mobile user u k ;

[0124] Under the LoS channel, the path loss between UAV A m and mobile user u k can be expressed as

[0125]

[0126] where, represents the distance between UAV A m and mobile user u k , f c represents the carrier frequency, c represents the speed of light, and η LoS represents the additional path loss corresponding to the LoS channel, which is a constant value depending on the environment;

[0127] Similarly, under the NLoS channel, the communication probability between UAV A m and mobile user u kThe communication probability between

[0128]

[0129] In the NLoS channel, the communication path loss is expressed as

[0130]

[0131] where η NLoS represents the corresponding additional path loss in the NLoS channel, which is a constant value depending on the environment;

[0132] In summary, for UAV A m and the mobile user u k the path loss between them can be expressed as

[0133]

[0134] In summary, for UAV A m and the mobile user u k the transmission rate can be expressed as

[0135]

[0136] where B represents the system transmission bandwidth, P represents the average transmit power of the user, and σ 2 represents the Gaussian white noise.

[0137] To ensure that communication can be established between the UAV and the user, the minimum channel gain g0 is used to represent the condition for the channel to be communicable. When the condition

[0138]

[0139] is satisfied, communication can be established between UAV A m and the mobile user u k

[0140] Under the condition that the horizontal and vertical positions of the UAV are restricted, the present invention aims to solve the problem of maximizing the transmission rate between the UAV and the mobile user. At the same time, to ensure the utilization rate of resources, while maximizing the transmission rate between the UAV and the mobile user, the number of UAVs is reduced as much as possible to maximize the average transmission rate of the system.

[0141] Based on the discussion of the model, the optimization problem of the present invention is modeled as follows

[0142]

[0143] s.t. X min ≤x m ≤X max (98)​

[0144] Y min ≤y m ≤Y max (99)

[0145] H min ≤h≤H max (100)

[0146]

[0147] The solution of the UAV position is a non-convex problem, and it is difficult to obtain the optimal solution using traditional algorithms. Therefore, the present invention proposes a scheme of dense boundary priority service: First, the original optimization problem is decomposed into the optimization of the vertical position and the optimization of the horizontal position. Using the KKT conditions, the maximum coverage area of the UAV in the current environment and its corresponding vertical height are solved. Then, fixing the vertical coordinate, the overall area is divided into multiple regions, and the horizontal position is solved for the UAV deployed in each region, so that the average transmission rate between the UAV and the mobile user in the current environment is maximized.

[0148] 1. Maximum service radius and sub-problem of vertical problem:

[0149] It can be deduced from the minimum channel gain that can communicate between the UAV and the mobile user:

[0150] PL m,k ≤ -10log(g0) (102)

[0151] Thus, the maximum service radius r of the UAV and the corresponding height h are obtained.

[0152] In summary, the sub-problem of obtaining the vertical position of the UAV and the maximum service radius can be expressed as:

[0153]

[0154] s.t. H min ≤h≤H max (104)

[0155]

[0156] This problem is a non-convex problem. By solving it using the KKT conditions, a sub-optimal solution is obtained.

[0157] First, use the trigonometric relationship h = rtanθ m,n , to convert the maximum service radius r of the UAV and its corresponding height h into the relationship between the maximum service radius r of the UAV and the elevation angle θ m,n between them, thus transforming the problem into:

[0158]

[0159] such that PL m,k + 10log(g0) ≤ 0 (107)

[0160] H min -rtanθ m,k ≤ 0 (108)

[0161] rtanθ m,k -H max ≤ 0 (109)

[0162] According to the above optimization problem, the Lagrangian function can be expressed as:

[0163]

[0164] where λ1, λ2, λ3 are Lagrange multipliers.

[0165] According to the KKT conditions, the parameters should satisfy

[0166]

[0167] λ2(H min -rtanθ m,k ) = 0 (113)

[0168] λ3(rtanθ m,k -H max ) = 0 (114)

[0169]

[0170] H min -rtanθ m,k ≤ 0 (117)

[0171] rtanθ m,k -H max ≤ 0 (118)

[0172] λ1 ≥ 0, λ2 ≥ 0, λ3 ≥ 0 (119)

[0173] In summary, it can be divided into three cases to derive the relevant expressions of r and θ m,k Related expressions:

[0174]

[0175] where θ m,k represents the angle between the drone A m and the mobile user u k and

[0176]

[0177] If λ2 > 0 and λ3 > 0, then h = H min = H max , which does not conform to the actual situation. Therefore, for the above three cases, under the condition of satisfying the λ2 and λ3 limit conditions, the corresponding r and θ are solved m,k And compare them, and the solution that makes the service radius r the largest is the final solution

[0178] 2. Region division and horizontal position sub - problem:

[0179] By solving sub - problem 1, the maximum service radius of the UAV and the corresponding height in the current environment can be obtained. At this time, the original problem can be simplified as:

[0180]

[0181] s.t. X min ≤ x m ≤ X max (122)

[0182] Y min ≤ y m ≤ Y max (123)

[0183]

[0184] For a given area, to solve the problem of the deployment of multiple UAVs, through an algorithm based on dense - boundary - first service, the overall area is divided into multiple partial areas, and the deployment problem of a single UAV is solved within the divided areas. For the partial area, equation (35) can be expressed as:

[0185]

[0186] According to equation (9), it can be known that R m,k increases as increases. So for equation (39), it is equivalent to finding

[0187]

[0188] For equation (41), in a general urban environment, η LoS < η NLoS . Therefore, when h is determined, as L=(x m - a k ) 2 +(y m - b k ) 2decreases with the increase of, so solving Equation (39) can be transformed into solving:

[0189]

[0190] That is, to solve for the position within the given area that can minimize the sum of the distances from the UAV to all users, which is the position where the UAV is deployed.

[0191] By solving the Hessian matrix of Equation (42):

[0192]

[0193] It can be obtained that a1 = 2, a1a4 - a2a3 = 4, so Problem (42) is a convex problem and can be solved by optimization tools such as CVX.

[0194] It is only necessary to obtain the horizontal coordinate of the UAV to maximize the system average transmission rate. The present invention adopts an algorithm based on dense boundary priority service to solve the problem of maximizing the system average transmission rate under the constraints of the UAV position and the minimum channel gain that can communicate.

[0195] First, in order to ensure that the UAV can cover more users and thus reduce the number of UAVs, define the boundary users as the current minimum and maximum a k , the minimum and maximum b k , the users at the four positions, and the corresponding positions are respectively

[0196] Then, in the area with a radius of 2r centered on the above four users, select the area that covers the most users as the area where the UAV will be deployed;

[0197] Finally, for the selected area, optimization tools such as CVX can be used for solution.

[0198] In Figure 3 it can be seen that for areas with dense users, UAVs are preferentially deployed. To ensure that the UAV can serve all users, for a user whose distance from other users is greater than 2r, a single UAV will be deployed separately. Figure 4Comparison chart of the number of drones required by the K-means algorithm and the DBPS algorithm (Dense Boundary Priority Service Method) when the number and distribution of users are the same in a 5000m×5000m area. When the number of users is small, the distribution is relatively loose, and the number of drones required by the K-means algorithm and the DBPS algorithm is almost the same. As the number of users increases, the number of drones required also increases. However, since the K-means algorithm is not only related to the user distribution but also depends on the selection of the clustering center and the number of clusters, the setting of the initial value has a greater impact on the algorithm performance. For the DBPS algorithm, following the principle of giving priority to the density of boundary users, it preferentially covers users in dense areas and is only related to the user distribution, reducing uncertain factors. When the number of users increases to a certain extent and the required drones can cover the entire area, as the number of users increases, no more drones will be deployed. Figure 5 Comparison chart of the system average transmission rates of the K-means algorithm and the DBPS algorithm when the number and distribution of users are the same in a 5000m×5000m area. As can be seen from the figure, when the number of users is small, Figure 4 it can be known that the number of drones is almost the same. Under the same user distribution, the system average transmission rates are at the same level. As the number of users increases, the number of drones required also increases, and the impact on the system average transmission rate also increases. From Figure 4 、 Figure 5 it can be seen that as the number of users increases, compared with the K-means algorithm, the DBPS algorithm can serve users in the area with a small number of drones, resulting in a better system average transmission rate than the K-means algorithm.

[0199] Based on this embodiment, the present invention also provides a drone deployment system based on dense boundary priority service, including:

[0200] An initialization module for obtaining relevant information in the current scenario, including ground user location information, the corresponding urban model information in the current environment, and the minimum channel gain that can communicate in the current environment;

[0201] An optimization model establishment module for establishing a system optimization model with the drone position and the communicable channel gain as constraints and maximizing the average transmission rate as the goal according to the overall optimization goal;

[0202] An optimization problem decomposition module for decomposing the optimization problem into a maximum service radius and vertical position sub-problem and a region division and horizontal position sub-problem;

[0203] A vertical position solving module for solving the maximum service radius and vertical position sub-problem based on the KKT condition;

[0204] The horizontal position solving module is used to solve the UAV horizontal position sub-problem based on the area division method of dense boundary priority service.

[0205] The present invention optimizes the three-dimensional positions of multiple UAVs in a UAV-assisted mobile edge network, achieving the goal of maximizing the system average transmission rate under the condition of limited UAV resources; the UAV deployment method implemented by the present invention only depends on the current urban model and user distribution, and this parameter can be obtained in the initialization stage. Moreover, compared with the traditional heuristic method, the method proposed by the present invention does not depend on the selection of the initial solution and can find the UAV deployment positions that can obtain a relatively large system average transmission rate.

[0206] The above is only a detailed description of the preferred embodiments and principles of the present invention. For those of ordinary skill in the art, according to the idea provided by the present invention, there will be changes in the specific implementation manners, and these changes should also be regarded as the protection scope of the present invention.

Claims

1. A method for deploying unmanned aerial vehicles based on dense boundary priority service, characterized in that It includes the following steps: S1, Initialization stage: Obtain relevant information in the current scenario, including ground user location information, the corresponding urban model information in the current environment, and the minimum channel gain that can communicate in the current environment; S2, Establish an optimization model: According to the overall optimization goal, establish a system optimization model with the drone position and the communicable channel gain as constraints and the maximization of the average transmission rate as the goal; S3, Decompose the optimization problem into a maximum service radius and vertical position sub-problem and a region division and horizontal position sub-problem; S4, Solve the maximum service radius and vertical position sub-problem based on the KKT conditions; S5, Solve the drone horizontal position sub-problem based on the region division method of dense boundary priority service; Step S1 includes the following steps: S11, define the set of drones as \(A = \{A_1, A_2, \ldots, A m , \ldots, A M \}\), and the set of users as \(U = \{u_1, u_2, \ldots, u k , \ldots, u K \}; assume that all drones have the same vertical height, and use a three-dimensional Cartesian coordinate system to represent the coordinates of the \(m\)-th drone as \((x m , y m , h)\), and the coordinates of the \(k\)-th user as \((a k , b k , 0)\), and assume that all user positions are fixed; there are the following restrictions on the drone positions: X min ≤x m ≤X max (1) Y min ≤ y m ≤ Y max (2) H min ≤h≤H max (3) where X min and X max and Y min and Y max and H min and H max respectively represent the minimum and maximum ranges of the horizontal and vertical positions of the drone; S12. Establish an LoS channel model in the urban environment. Under the LoS channel, the communication probability between UAV A m and the mobile user u k is as follows: where a and b are constant values depending on the current environment, denotes the UAV A m and the mobile user u k the included angle therebetween; Under the LoS channel, UAV A m and mobile user u k The path loss between them is expressed as: Among them, represents the drone A m and the mobile user u k The distance between, f c represents the carrier frequency, c represents the speed of light, η LoS represents the additional path loss corresponding to the LoS channel, which is a constant value depending on the environment; S13. In the NLoS channel, the communication probability between UAV A m and mobile user u k is as follows: In the NLoS channel, the communication path loss is expressed as: Among them, η NLoS represents the additional path loss corresponding to the NLoS channel and is a constant value depending on the environment; It is obtained that for drone A m and mobile user u k the path loss between them is expressed as: It is concluded that the drone A m and the mobile user u k The transmission rate of is expressed as Among them, B represents the system transmission bandwidth, P represents the average transmission power of users, and σ 2 represents Gaussian white noise; Use the minimum channel gain g0 to represent the condition for channel communication. When the condition is satisfied When, the UAV A m and the mobile user u k can establish communication; Step S2 includes the following steps: S21, The system optimization model is modeled as follows: s.t. X min ≤x m ≤X max (12) Y min ≤ y m ≤ Y max (13) H min ≤h≤H max (14) Step S4 includes the following steps: S41, Deduce from the minimum channel gain that can communicate between the drone and the mobile user: PL m,k ≤ -10 log(g0) (16) Furthermore, obtain the maximum service radius r of the drone and the corresponding height h; It is concluded that the sub-problem of the drone vertical position and the maximum service radius is expressed as: such that H min ≤h≤H max (18) S42. Using the trigonometric relationship h = rtanθ m,k , convert the maximum service radius r of the UAV and the corresponding height h into the relationship between the maximum service radius r of the UAV and the elevation angle θ m,k between them, and transform the problem into: s.t. PL m,k +10log(g0)≤0 (21) H min -rtanθ m,k ≤0 (22) rtanθ m,k -H max ≤0 (23) According to the above optimization problem, it is expressed by the Lagrangian function as: Among them, λ1, λ2, and λ3 are Lagrange multipliers; S43, According to the KKT conditions, the parameters satisfy: λ2(H min -rtanθ m,k ) = 0 (27) λ3(rtanθ m,k -H max ) = 0 (28) H min -rtanθ m,k ≤0 (31) rtanθ m,k -H max ≤0 (32) λ1≥0, λ2≥0, λ3≥0 (33) In summary, there are three cases, and \(r\) and \(\theta\) are derived m,k Related expressions: where θ m,k represents the angle m between the drone A k and the mobile user u For the above three cases, when the λ2 and λ3 constraint conditions are satisfied, the corresponding r and θ are solved m,k And a comparison is made, and the solution with the largest service radius r is the final solution; Step S5 includes the following steps: S51, Through the solution of step S4, obtain the maximum service radius and the corresponding height of the drone in the current environment. At this time, the optimization model is simplified to: such that X min ≤ x m ≤ X max (36) Y min ≤y m ≤Y max (37) For a given area, solve the problem of the deployment of multiple drones. Through the algorithm based on dense boundary priority service, divide the overall area into multiple partial areas, and solve the deployment problem of a single drone within the divided area; For the partial area, formula (35) is expressed as: According to Equation (9), R m,k increases as it increases. For Equation (39), it is equivalent to finding For Equation (41), in a general urban environment, η LoS < η NLoS , it is obtained that when h is determined, as L = (x m - a k ) 2 + (y m - b k ) 2 increases, it decreases, and it is obtained that solving Equation (39) is transformed into solving: That is, solve the position within the given area that minimizes the sum of the distances from the drone to all users, which is the position of the drone deployment, By solving the Hessian matrix corresponding to formula (42): It is obtained that a1 = 2, a1a4 - a2a3 = 4, and solve it through the CVX optimization tool.

2. The drone deployment system based on dense boundary priority service is used to implement the drone deployment method based on dense boundary priority service described in claim 1, and is characterized in that, The drone deployment system based on dense boundary priority service includes: An initialization module, used to obtain relevant information in the current scenario, including ground user location information, the corresponding urban model information in the current environment, and the minimum channel gain that can communicate in the current environment; An optimization model establishment module, used to establish a system optimization model with the drone position and the communicable channel gain as constraints and the maximization of the average transmission rate as the goal according to the overall optimization goal; An optimization problem decomposition module, used to decompose the optimization problem into a maximum service radius and vertical position sub-problem and a region division and horizontal position sub-problem; A vertical position solution module, used to solve the maximum service radius and vertical position sub-problem based on the KKT conditions; A horizontal position solution module, used to solve the drone horizontal position sub-problem based on the region division method of dense boundary priority service.

Images