Method and device for optimizing the coordinated deployment of multiple UAV base stations under building obstacles

Abstract

An embodiment of the present invention provides a method and device for optimizing the coordinated deployment of multiple UAV base stations under architectural obstacles. The method includes: randomly assigning position coordinates to all UAVs in a candidate space multiple times to obtain multiple solutions; for each solution, calculating the number of users that can be served by all UAVs in the solution as the fitness value corresponding to the solution; sorting all solutions in ascending order of their corresponding fitness values ​​to obtain a solution ranking, taking the solution with the largest fitness value in the solution ranking as the global optimal solution, and taking the corresponding fitness value as the global optimal fitness value; based on an improved biogeographic optimization algorithm, cyclically iteratively updating the multiple solutions, sorting the multiple solutions according to the updated fitness values, updating the global optimal solution and the global optimal fitness value according to the obtained ranking, and stopping the iteration until the number of iterations reaches a maximum number of iterations or the number of consecutive no-improvement times reaches a maximum number of consecutive no-improvement times, and using the global optimal solution for the coordinated deployment of multiple UAV base stations.

Description

Technical Field

[0001] The present invention relates to the field of drone technology, and in particular to a method and device for optimizing the coordinated deployment of multiple drone base stations under building obstacles. Background Art

[0002] The rapid development of IoT technology has made it a critical infrastructure for modern society. Its widespread application in daily life and industry has highlighted the urgent need for robust and flexible temporary network systems. However, traditional network infrastructure is unable to cope with the sudden network demands in emergency communication scenarios such as disaster response and military operations, highlighting the need for innovative solutions to meet temporary network needs.

[0003] Drone base stations, with their cost-effectiveness and rapid response, offer an innovative solution for emergency communications. Their low cost makes them readily available for large-scale deployment during critical periods like disaster response. Their agility allows for rapid deployment to remote areas, providing users with immediate network access. Furthermore, drone base stations can be deployed at low altitudes, easily establishing line-of-sight with users on the ground and providing high-quality communications. These advantages make them ideal for restoring or enhancing network coverage when ground infrastructure is damaged or overloaded.

[0004] Determining the optimal deployment locations for drone base stations is a critical and complex challenge. Buildings, as major obstacles in urban environments, not only occupy the space available for drone deployment but also act as a key barrier to communication signal propagation between drone base stations and users. Therefore, the type of communication link between the drone base station and users needs to be determined based on whether the building is blocking the communication link: a non-line-of-sight link if blocked, and a line-of-sight link otherwise. Deploying a limited number of drone base stations to provide maximum coverage for randomly distributed users is essential in resource-constrained environments. By ensuring that drone base stations maintain dual connectivity with other drone base stations, a robust drone network can be achieved, effectively preventing potential threats from disrupting the drone network.

[0005] Many scholars have studied the problem of multi-UAV base station deployment. However, most current research on UAV base station deployment relies on statistical channel models with a standard building distribution, which cannot capture the communication link congestion caused by the actual building distribution in different environments. More importantly, the optimization problem of UAV deployment is a cutting-edge application of the facility site selection problem. It is usually based on the Maximum Covering Location Problem (MCLP) model, which integrates the characteristics of UAV communication for modeling. Key constraints include deployment height, service capacity, user allocation, and connectivity between UAVs. However, maintaining dual connections for each UAV to form a robust UAV network has not been sufficiently studied. Therefore, there is an urgent need to study new UAV deployment methods that consider building obstacles and robust UAV networks.

[0006] In the process of implementing the present invention, the applicant discovered that the prior art has at least the following problems:

[0007] How to deploy drone network base stations in an environment with multiple building obstacles to maximize the coverage of the drone network. Summary of the Invention

[0008] The embodiments of the present invention provide a method and device for optimizing the coordinated deployment of multiple drone base stations in the presence of building obstacles, which solves the problem of how to deploy drone network base stations in an environment with multiple building obstacles to achieve maximum coverage of the drone network.

[0009] To achieve the above objectives, on the one hand, an embodiment of the present invention provides a method for optimizing the coordinated deployment of multiple UAV base stations in the presence of building obstacles, comprising:

[0010] Randomly assigning position coordinates to all drones in a candidate space multiple times to obtain multiple solutions, each solution including the position coordinates of all drones; the candidate space is a space in the environment where the drones are deployed that is not occupied by buildings;

[0011] For each solution, the number of users that can be served by all the drones in the solution is calculated based on the position coordinates of all the drones in the solution and the preset user coordinates of multiple users, and the number of users that can be served by all the drones in the solution is used as the fitness value corresponding to the solution;

[0012] All solutions are sorted in ascending order according to their corresponding fitness values ​​to obtain a solution ranking, the solution with the largest fitness value in the solution ranking is taken as the global optimal solution, and the fitness value corresponding to the global optimal solution is taken as the global optimal fitness value;

[0013] Based on the improved biogeographic optimization algorithm, the multiple solutions are iteratively updated in a loop, and the multiple solutions are sorted according to the fitness values ​​of the multiple solutions updated in each iteration. The global optimal solution and the global optimal fitness value are updated according to the obtained sorting. The iteration is stopped until the number of loop iterations reaches a preset maximum number of iterations or the number of consecutive no improvements reaches a preset maximum number of consecutive no improvements. The global optimal solution is used for the coordinated deployment of multiple UAV base stations, and the obtained global optimal fitness value is the number of users that can be served after the coordinated deployment of multiple UAV base stations according to the obtained global optimal solution.

[0014] On the other hand, an embodiment of the present invention provides a device for optimizing the coordinated deployment of multiple UAV base stations in the presence of building obstacles, characterized by comprising:

[0015] An initial solution determination module is configured to randomly assign position coordinates to all drones in a candidate space multiple times to obtain multiple solutions, each solution including the position coordinates of all drones; the candidate space is a space in the environment where the drones are deployed that is not occupied by buildings;

[0016] a fitness value determination module, configured to calculate, for each solution, the number of users that can be served by all the drones in the solution based on the position coordinates of all the drones in the solution and the preset user coordinates of multiple users, and use the number of users that can be served by all the drones in the solution as the fitness value corresponding to the solution;

[0017] a global optimal determination unit, configured to sort all solutions in ascending order of their corresponding fitness values ​​to obtain a solution ranking, taking the solution with the largest fitness value in the solution ranking as the global optimal solution, and taking the fitness value corresponding to the global optimal solution as the global optimal fitness value;

[0018] The global optimal iteration unit is used to perform cyclic iterative updates on the multiple solutions based on the improved biogeographic optimization algorithm, and sort the multiple solutions according to the fitness values ​​of the multiple solutions updated in each iteration according to the updated fitness values, and update the global optimal solution and the global optimal fitness value according to the obtained sorting, until the number of cyclic iterations reaches a preset maximum number of iterations or the number of consecutive no improvements reaches a preset maximum number of consecutive no improvements, and the iteration is stopped. The global optimal solution is used for the coordinated deployment of multiple UAV base stations, and the obtained global optimal fitness value is the number of users that can be served after the coordinated deployment of multiple UAV base stations according to the obtained global optimal solution.

[0019] The above technical solution has the following beneficial effects: by iteratively updating the global optimal solution for drone deployment using an improved biogeographic optimization algorithm, it is possible to achieve effective coordinated deployment of multiple drone base stations, taking into account architectural obstacles. This solution not only produces optimal results but also exhibits faster convergence, making it more suitable for solving the coordinated deployment optimization problem of multiple drone base stations. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0021] Figure 1 This is a flow chart of a method for optimizing the coordinated deployment of multiple UAV base stations under building obstacles according to one embodiment of the present invention;

[0022] Figure 2 This is an architectural diagram of a multi-UAV base station collaborative deployment optimization device under building obstacles, one of the embodiments of the present invention;

[0023] Figure 3 This is a schematic diagram of drone base station deployment considering building obstacles according to one embodiment of the present invention;

[0024] Figure 4 This is a schematic diagram of the Los link and NLoS link between the drone base station and the user according to one embodiment of the present invention;

[0025] Figure 5 is a schematic diagram of a straddling method according to one embodiment of the present invention;

[0026] Figure 6 It is a two-dimensional top view of the Beichen Delta region according to one embodiment of the present invention;

[0027] Figure 7 This is a schematic diagram of the simulated distribution of buildings and users in the Beichen Delta region according to one embodiment of the present invention;

[0028] Figure 8 It is the three-dimensional optimal deployment result solved by the RD+BBO method of one embodiment of the present invention;

[0029] Figure 9 It is the two-dimensional optimal deployment result solved by the RD+BBO method of one embodiment of the present invention;

[0030] Figure 10 This is the three-dimensional optimal deployment result obtained by the KMC+BBO method according to one embodiment of the present invention;

[0031] Figure 11 This is the two-dimensional optimal deployment result obtained by the KMC+BBO method according to one embodiment of the present invention;

[0032] Figure 12 is the convergence curve of the RD+BBO and KMC+BBO methods according to one embodiment of the present invention;

[0033] Figure 13 FIG2 is a schematic diagram of a drone providing wireless services according to one embodiment of the present invention. DETAILED DESCRIPTION

[0034] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0035] With the rapid development of the Internet of Things (IoT) and its increasing application in social life and industrial production, the robustness and flexibility of network infrastructure are being increasingly demanded. In particular, in emergency situations such as disaster response, the limitations of traditional networks are becoming increasingly apparent, and innovative solutions are urgently needed to meet temporary network needs. Drones (drones) equipped with mobile base stations are becoming an effective means of providing network services in emergency situations due to their low cost and rapid deployment. A key challenge is how to deploy limited drone base stations while maximizing coverage for users on the ground while accounting for building obstructions and ensuring the robustness of the drone network. Considering the obstruction of wireless signals by urban buildings, a binary channel model is employed to classify the communication links between users and drone base stations as either line-of-sight or non-line-of-sight. The problem of optimizing the coordinated deployment of multiple drone base stations is proposed, taking into account capacity, coverage, and dual connectivity constraints. Based on the characteristics of drone swarms, an improved biogeographic optimization algorithm is designed to solve this problem. Finally, a real-world case study is applied to validate the effectiveness of the proposed method.

[0036] The embodiment of the present invention solves a key challenge, namely how to deploy a limited number of drone base stations to achieve maximum coverage under the conditions of building obstacles and drone network robustness. Depending on whether there are obstacles blocking the drone and the user, the communication link can be divided into line of sight (LoS) and non-line of sight (NLoS). Based on MCLP, the collaborative deployment optimization problem of multiple drone base stations is established, taking into account user coverage and drone dual connection constraints. Utilizing the intelligence of drone clusters, an improved biogeographic optimization algorithm is designed, and a real case is applied to verify its effectiveness. In this patent, drones are often referred to as drone base stations.

[0037] On the one hand, if Figure 1 As shown, an embodiment of the present invention provides a method for optimizing the coordinated deployment of multiple UAV base stations under building obstacles, including:

[0038] Step S10, randomly assigning position coordinates to all drones in a candidate space multiple times to obtain multiple solutions, each solution including the position coordinates of all drones; the candidate space is a space in the environment where the drones are deployed that is not occupied by buildings;

[0039] Step S11: For each solution, based on the position coordinates of all drones in the solution and the preset user coordinates of multiple users, calculate the number of users that can be served by all drones in the solution, and use the number of users that can be served by all drones in the solution as the fitness value corresponding to the solution;

[0040] Step S12, sorting all solutions in ascending order of their corresponding fitness values ​​to obtain a solution ranking, taking the solution with the largest fitness value in the solution ranking as the global optimal solution, and taking the fitness value corresponding to the global optimal solution as the global optimal fitness value;

[0041] Step S13: Based on the improved biogeographic optimization algorithm, the multiple solutions are iteratively updated in a loop, and the multiple solutions are sorted according to the fitness values ​​of the multiple solutions updated in each iteration. The global optimal solution and the global optimal fitness value are updated according to the obtained sorting. The iteration is stopped until the number of loop iterations reaches a preset maximum number of iterations or the number of consecutive no-improvement times reaches a preset maximum number of consecutive no-improvement times. The global optimal solution is used for the coordinated deployment of multiple UAV base stations. The obtained global optimal fitness value is the number of users that can be served after the coordinated deployment of multiple UAV base stations according to the obtained global optimal solution.

[0042] The embodiments of the present invention have the following technical advantages: By iteratively updating the global optimal solution for the deployment of drones (drones equipped with base stations for wireless communication with users) using an improved biogeographic optimization algorithm, the solution can achieve efficient coordinated deployment of multiple drone base stations, taking into account architectural obstacles. This solution not only produces optimal results but also exhibits faster convergence, making it more suitable for solving the coordinated deployment optimization problem of multiple drone base stations.

[0043] Furthermore, the multiple random assignments of position coordinates for all drones in the candidate space are performed to obtain multiple solutions, each solution including the position coordinates of all drones, including:

[0044] Acquiring spatial information of an environmental space, rasterizing the environmental space into a plurality of continuously arranged cubic grids, and obtaining a grid set of buildings within the environmental space and a grid set of candidate spaces; the grid set comprising a plurality of cubic grids; the spatial information of the environmental space comprising: the length, width, and height of the environmental space, and spatial information of one or more buildings within the environmental space; the spatial information of the buildings comprising the location coordinates, length, width, and height of the buildings;

[0045] Generating the three-dimensional position coordinates of all drones in each solution in a random manner within the grid set of the candidate space; and / or,

[0046] The k-means clustering method is used to generate the horizontal coordinates of all drones in the candidate space, and then the height coordinates of all drones are randomly determined within the preset candidate space height range; the position coordinates of a drone are the center coordinates of the cubic grid occupied by the drone; the position coordinates include horizontal coordinates and height coordinates.

[0047] Among them, the k-means clustering method is used to generate the horizontal coordinates of all drones in the candidate space, including:

[0048] Performing k-means clustering on the multiple users according to horizontal coordinates in the preset user coordinates to obtain horizontal coordinates of multiple cluster centers, and using the horizontal coordinates of the multiple cluster centers as the horizontal coordinates of the multiple drones;

[0049] Horizontal coordinates are generated based on user clusters, with cluster centers used as the horizontal coordinates of the drone. Compared to random methods that randomly generate initial solutions across the entire space, k-means clustering limits the spatial range of the initial solution generation, theoretically closer to optimal coverage. Therefore, the initial solution generated based on k-means clustering is of higher quality and closer to the global optimum.

[0050] Randomly determine the height coordinates of all drones within the preset candidate space height range, including:

[0051] For each drone, the altitude range of the drone can be determined based on the path loss between the drone and the user (under LoS and nLoS links), provided that the path loss is no greater than a preset loss threshold and the maximum radius that the drone's transceiver power can cover, and the altitude coordinates of the drone can be randomly assigned within the obtained altitude range;

[0052] Randomly determine the height coordinates of all drones within the preset candidate space height range, including:

[0053] For each drone, the optimal deployment height corresponding to the maximum coverage radius of the drone can also be determined based on the probabilistic statistical channel model. With the optimal deployment height as the center, the upper and lower limits of the randomly selected range of the drone's height coordinates are determined. The upper limit is equal to the optimal deployment height plus a preset upward offset value, and the lower limit is equal to the optimal deployment height minus a preset downward offset value. The probabilistic statistical channel model refers to the path loss between the drone and the user being composed of LoS and nLoS links with a certain fixed probability. The path loss depends on the environmental parameters related to the building, the distance between the drone and the user, and the elevation angle. Users usually have minimum requirements for service quality, which can be expressed by path loss. If a certain path loss threshold is given, a nonlinear implicit function of the service height with respect to the coverage radius can be obtained by simplification, and the optimal service height (optimal deployment height) corresponding to the maximum coverage radius can be obtained by taking the partial derivative. The connection between the drone and the user is transmitted through the air-to-ground channel. The air-to-ground channel model is composed of LoS and nLoS links with a certain probability, so the path loss from drone k to user i can be expressed as formula (1):

[0054] L(h,d ik )=L LoS ×P(L LoS ,θ ik )+L nLoS ×P(L nLoS ,θ ik ) (1)

[0055] The probability of a LoS link depends on the building density, the proportion of building area, and the location of the drone and the user, and can be expressed as formula (2):

[0056]

[0057] Where a and b are constant coefficients determined by the environment (suburbs, cities, dense cities, high-rise buildings, etc.), is the elevation angle from drone k to user i, h is the height of the drone, d ik is the distance from user i to the center of the coverage circle of drone k, such as Figure 13As shown. The probability of nLoS link is formula (3):

[0058] P(nLoS,θ ik )=1-P(LoS,θ ik ) (3).

[0059] The path loss for LoS and nLoS links can be expressed as Equation (4) and Equation (5), respectively:

[0060]

[0061] where ηLoS and ηnLoS are the average additional losses of LoS and nLoS respectively, and f c is the carrier frequency of the air-to-ground channel and c is the speed of light.

[0062] Then, formula (1) can be simplified to formula (6):

[0063]

[0064] Given a preset path loss threshold L(h,d ik ), formula (6) is the nonlinear implicit function of the height of the drone with respect to the distance from the user to the center of the unmanned k coverage circle. By taking the partial derivative of formula (6), we can find the inflection point of the function, that is, the optimal height (optimal deployment height) H, and d ik The maximum value is the maximum coverage radius R.

[0065] Furthermore, for each solution, the number of users that can be served by all the drones in the solution is calculated based on the position coordinates of all the drones in the solution and the preset user coordinates of multiple users, and the number of users that can be served by all the drones in the solution is used as the fitness value corresponding to the solution, including:

[0066] For each solution, calculate whether there is a building obstruction between each user and each drone in the solution based on the preset user coordinates of each user, the location coordinates of each drone in the solution, and the spatial information of all buildings in the environment space; the spatial information of the building includes the location coordinates, length, width, and height of the building;

[0067] For each user, for links between drones with building obstructions, calculate the NLoS received power (NROP) obtained by the user from the drones with building obstructions. For links between drones with no building obstructions, calculate the LoS received power obtained by the user from the drones without building obstructions.

[0068] For each user, the user is assigned to a drone corresponding to the maximum value of all the NLoS reception powers and LoS reception powers of the user;

[0069] For each solution, the total number of users served by all drones in the solution is counted as the fitness value corresponding to the solution.

[0070] Furthermore, for each solution, based on the preset user coordinates of each user, the position coordinates of each drone in the solution, and the spatial information of all buildings in the environment space, calculating whether there is a building obstruction between each user and each drone in the solution includes:

[0071] The spanning method in computational geometry is used to determine whether there is a building obstruction between each drone and the user;

[0072] The straddle method uses the mathematical properties of vector cross product and dot product to determine whether two line segments intersect. Specifically, there are two necessary and sufficient conditions for the intersection of line segments P1P2 and Q1Q2:

[0073] Condition one:

[0074] in, The vector and vector The direction of the cross product vector, The vector and vector The direction of the vector of the cross product result is that the line segment P1P2 will cross the line segment Q1Q2 only when the dot product of the two result vectors is greater than 0;

[0075] Condition two:

[0076] Condition 2 is that line segment Q1Q2 is distributed on both sides of line segment P1P2;

[0077] If these two conditions are met at the same time, the line segment P1P2 intersects with the line segment Q1Q2, and it can be concluded that there is a building blocking the drone and the user.

[0078] In specific applications, the line segment P1P2 refers to the line connecting the user and the drone's projection on the bottom surface, and the line segment Q1Q2 refers to the four sides of the bottom surface of the building.

[0079] Furthermore, for a link between a drone and the user in the case of an obstacle of a building blocking the drone, calculating the NLoS received power obtained by the user from the drone with the building blocking the drone, and for a link between a drone and the user without a building blocking the drone, calculating the LoS received power obtained by the user from the drone without a building blocking the drone, specifically includes:

[0080] Based on the binary channel model, the power loss of air-to-ground communication is calculated by the following formula:

[0081] PL LoS =20lg(d ij )+20lg(f c )+η LoS (9)

[0082] PL NLoS =20lg(d ij )+20lg(f c )+η NLoS (10)

[0083] Among them, f c is the carrier frequency, d ij is the distance between user i and drone j, η LoS and η NLoS represents the additional path power loss under link state LoS and NLoS respectively; η LoS =92.4,η NLoS =92.4+L s , additional random path loss under building occlusion conditions normarnd(μ,σ) represents normal distribution, θ ij is the elevation angle of user i to drone j, g μ 、g σ 、h μ 、h σ 、i μ 、i σ is an empirical parameter;

[0084] The NLoS received power and LoS received power are calculated using the following formula:

[0085] P LoS =P t -PL LoS (11)

[0086] P NLoS =P t -PL NLoS (12)

[0087] Among them, P tTransmit power to the drone, PL LoS and PL NLoS represent the path loss under LoS and NLoS links respectively, P LoS and P NLoS They represent the user's LoS received power and NLoS received power when the link status is LoS and NLoS, respectively.

[0088] Furthermore, the improved biogeographic optimization algorithm is based on a cyclic iterative update of the multiple solutions, and the multiple solutions are sorted according to the fitness values ​​of the multiple solutions updated in each iteration according to the updated fitness values, and the global optimal solution and the global optimal fitness value are updated according to the obtained sorting. The iteration is stopped until the number of cyclic iterations reaches a preset maximum number of iterations or the number of consecutive no-improvements reaches a preset maximum number of consecutive no-improvements. The obtained global optimal solution is used for the coordinated deployment of multiple UAV base stations. The obtained global optimal fitness value is the number of users that can be served after the coordinated deployment of multiple UAV base stations according to the obtained global optimal solution, including:

[0089] Initialize the elite rate, determine the number of retained solutions and the number of updated solutions for each iteration based on the elite rate and the total number of preset solutions, and set the preset mutation probability p M ;

[0090] The following iterative loop is executed until the maximum number of iterations is reached, or the iteration is stopped when the number of consecutive no-improvement times reaches the maximum number of consecutive no-improvement times:

[0091] According to the sorting of all solutions, the solutions with high fitness values ​​and the number of retained solutions are taken as the solutions that remain unchanged among all solutions, and the remaining solutions with low fitness values ​​and the number of updated solutions are taken as the solutions that need to be updated among all solutions;

[0092] Performing a migration operation on each of the solutions that need to be updated, and updating the solutions that need to be updated;

[0093] Performing a mutation operation on the solution that needs to be updated after the migration operation is performed, so as to update the solution that needs to be updated;

[0094] All solutions are evaluated and modified based on connectivity constraints, where each UAV is constrained to have at least two neighboring UAVs, where the neighboring UAVs are UAVs whose distance is less than a preset neighbor distance threshold.

[0095] For each solution, the number of users that can be served by all the drones in the solution is calculated based on the position coordinates of all the drones in the solution and the preset user coordinates of multiple users, and the fitness value corresponding to the solution is updated using the number of users that can be served by all the drones in the solution;

[0096] Sorting all solutions in ascending order of their corresponding updated fitness values, updating the solution sorting, using the solution with the largest fitness value in the solution sorting to update the global optimal solution, and using the fitness value corresponding to the global optimal solution to update the global optimal fitness value;

[0097] When the iteration is stopped, the global optimal solution and the global optimal fitness value are output.

[0098] Among them, for each solution, the number of users that can be served by all drones in the solution is calculated according to the position coordinates of all drones in the solution and the preset user coordinates of multiple users, and the fitness value corresponding to the solution is updated using the number of users that can be served by all drones in the solution. Reference can be made to the aforementioned embodiment in which, for each solution, the number of users that can be served by all drones in the solution is calculated according to the position coordinates of all drones in the solution and the preset user coordinates of multiple users, and the number of users that can be served by all drones in the solution is used as an explanation of the fitness value corresponding to the solution.

[0099] Furthermore, performing a migration operation on each of the solutions that need to be updated to update the solutions that need to be updated includes:

[0100] According to the solution ranking of all solutions, set the immigration rate and emigration rate for all solutions, among which the solution with a large fitness value corresponds to a large immigration rate; the immigration rate and emigration rate are both less than or equal to 1, and the sum of the immigration rate and emigration rate of the same solution is 1;

[0101] For each solution that needs to be updated, a first random number corresponding to the solution is generated in the interval (0, 1). If the first random number corresponding to the solution is less than the immigration rate corresponding to the solution, the solution is used as the immigration solution, and the horizontal coordinates of a drone randomly selected from the outgoing solution are replaced with the horizontal coordinates of the drone randomly selected from the incoming solution. The outgoing solution is selected using a roulette algorithm from all solutions remaining after removing the incoming solution.

[0102] Furthermore, setting an in-migration rate and an out-migration rate for all solutions according to the solution sorting includes:

[0103] An ascending arithmetic sequence is generated between [0, 1], and terms in the arithmetic sequence are used as migration rates, wherein the number of terms in the arithmetic sequence is the same as the number of the multiple solutions;

[0104] According to the order of the solutions in the solution sorting and the terms in the arithmetic progression, the migration rate corresponding to each solution is obtained. The migration rate corresponding to each solution is equal to 1 minus the migration rate of the solution to one.

[0105] Furthermore, performing a mutation operation on the solution that needs to be updated after the migration operation is performed to update the solution that needs to be updated includes:

[0106] For each solution that needs to be updated, generate a second random number from the interval (0, 1), and if the second random number is less than or equal to the preset mutation probability, use the solution as a mutation solution;

[0107] For each drone in the mutation solution, the position coordinates of the drone are replaced by position coordinates randomly generated in the candidate space.

[0108] Furthermore, the evaluation and correction of all solutions based on connectivity constraints include:

[0109] For each solution, each drone in the solution is used as a target drone, and the distance between the target drone and other drones in the solution is calculated. If the obtained distance is less than a preset neighbor distance threshold, the other drone is added as a neighbor drone of the target drone;

[0110] Check if each target drone has at least two neighbor drones;

[0111] If it is found that the target drone has fewer than two neighboring drones, the target drone is adjusted toward the nearest non-neighboring drone by a moving distance, with the line connecting the target drone as the direction, to satisfy the connectivity constraint. The moving distance is the actual distance between the target drone and the non-neighboring drone minus a preset neighbor distance threshold.

[0112] On the other hand, Figure 2 As shown, an embodiment of the present invention provides a multi-UAV base station coordinated deployment optimization device under building obstacle conditions, comprising:

[0113] An initial solution determination unit 200 is configured to randomly assign position coordinates to all drones in a candidate space multiple times to obtain multiple solutions, each solution including the position coordinates of all drones; the candidate space is a space in the environment where the drones are deployed that is not occupied by buildings;

[0114] A fitness value determining unit 201 is configured to calculate, for each solution, the number of users that can be served by all the drones in the solution based on the position coordinates of all the drones in the solution and the preset user coordinates of multiple users, and use the number of users that can be served by all the drones in the solution as the fitness value corresponding to the solution;

[0115] A global optimal determination unit 202 is configured to sort all solutions in ascending order of their corresponding fitness values ​​to obtain a solution ranking, select the solution with the largest fitness value in the solution ranking as the global optimal solution, and select the fitness value corresponding to the global optimal solution as the global optimal fitness value;

[0116] The global optimal iteration unit 203 is used to perform cyclic iterative updates on the multiple solutions based on the improved biogeographic optimization algorithm, and sort the multiple solutions according to the fitness values ​​of the multiple solutions updated in each iteration according to the updated fitness values, and update the global optimal solution and the global optimal fitness value according to the obtained sorting. The iteration is stopped until the number of cyclic iterations reaches a preset maximum number of iterations or the number of consecutive no-improvement times reaches a preset maximum number of consecutive no-improvement times. The global optimal solution is used for the coordinated deployment of multiple UAV base stations, and the obtained global optimal fitness value is the number of users that can be served after the coordinated deployment of multiple UAV base stations according to the obtained global optimal solution.

[0117] Furthermore, the initial solution determining unit 200 includes:

[0118] An environmental space rasterization module is configured to obtain spatial information of an environmental space, rasterize the environmental space into a plurality of continuously arranged cubic grids, and obtain a grid set of buildings within the environmental space and a grid set of candidate spaces; the grid set includes a plurality of cubic grids; the spatial information of the environmental space includes: the length, width, and height of the environmental space, and spatial information of one or more buildings within the environmental space; the spatial information of the buildings includes the location coordinates, length, width, and height of the buildings;

[0119] A random position allocation module is used to randomly generate the three-dimensional position coordinates of all drones in each solution within the grid set of the candidate space; and / or,

[0120] The clustering and allocation position module is used to generate the horizontal coordinates of all drones in the candidate space using the k-means clustering method, and then randomly determine the height coordinates of all drones within the preset candidate space height range; wherein the position coordinates of a drone are the center coordinates of the cubic grid occupied by the drone; the position coordinates include horizontal coordinates and height coordinates.

[0121] The clustering allocation location module specifically includes:

[0122] Performing k-means clustering on the multiple users according to horizontal coordinates in the preset user coordinates to obtain horizontal coordinates of multiple cluster centers, and using the horizontal coordinates of the multiple cluster centers as the horizontal coordinates of the multiple drones;

[0123] Horizontal coordinates are generated based on user clusters, with cluster centers used as the horizontal coordinates of the drone. Compared to random methods that randomly generate initial solutions across the entire space, k-means clustering limits the spatial range of the initial solution generation, theoretically closer to optimal coverage. Therefore, the initial solution generated based on k-means clustering is of higher quality and closer to the global optimum.

[0124] Randomly determine the height coordinates of all drones within the preset candidate space height range, including:

[0125] For each drone, the altitude range of the drone can be determined based on the path loss between the drone and the user (under LoS and nLoS links), provided that the path loss is no greater than a preset loss threshold and the maximum radius that the drone's transceiver power can cover, and the altitude coordinates of the drone can be randomly assigned within the obtained altitude range;

[0126] Randomly determine the height coordinates of all drones within the preset candidate space height range, including:

[0127] For each drone, an optimal deployment altitude corresponding to the drone's maximum coverage radius can also be determined based on a probabilistic statistical channel model. With the optimal deployment altitude as the center, the upper and lower altitude limits of a randomly selected range for the drone's altitude coordinate are determined. The maximum coverage radius is determined based on the maximum radius that can be covered by the transmit / receive power of the drone's base station. The upper altitude limit is equal to the optimal deployment altitude plus a preset upward offset value, and the lower altitude limit is equal to the optimal deployment altitude minus a preset downward offset value. The probabilistic statistical channel model states that the path loss between the drone and the user consists of LoS and nLoS links, each with a fixed probability. This path loss depends on environmental parameters related to buildings, the distance between the drone and the user, and the elevation angle. Users typically have minimum quality of service requirements, which can be expressed as path loss. Given a certain path loss threshold, simplification yields a nonlinear implicit function of the service altitude with respect to the coverage radius. The optimal service altitude corresponding to the maximum coverage radius can be obtained by taking its partial derivative.

[0128] Furthermore, the fitness value determining unit 201 includes:

[0129] An occlusion determination module is configured to calculate, for each solution, whether there is a building occlusion between each user and each drone in the solution based on the preset user coordinates of each user, the location coordinates of each drone in the solution, and the spatial information of all buildings in the environment space; the spatial information of the building includes the location coordinates, length, width, and height of the building;

[0130] a power determination module configured to calculate, for each user, for links between drones that are blocked by buildings and the user, the NLoS received power obtained by the user from the drones blocked by buildings, and for links between drones that are not blocked by buildings and the user, the LoS received power obtained by the user from the drones not blocked by buildings;

[0131] A user allocation module is configured to allocate, for each user, a service to a drone corresponding to a maximum value of all NLoS received powers and LoS received powers of the user;

[0132] The fitness determination module is used to count the total number of users served by all drones in the solution for each solution as the fitness value corresponding to the solution.

[0133] Furthermore, the occlusion judgment module is configured as follows:

[0134] The spanning method in computational geometry is used to determine whether there is a building obstruction between each drone and the user;

[0135] The straddle method uses the mathematical properties of vector cross product and dot product to determine whether two line segments intersect. Specifically, there are two necessary and sufficient conditions for the intersection of line segments P1P2 and Q1Q2:

[0136] Condition 1 is formula (7), and condition 2 is formula (8). Condition 2 is the condition that the line segment Q1Q2 is distributed on both sides of the line segment P1P2;

[0137] If these two conditions are met at the same time, the line segment P1P2 intersects with the line segment Q1Q2, and it can be concluded that there is a building blocking the drone and the user.

[0138] Furthermore, the power determination module is configured as follows:

[0139] Based on the binary channel model, the power loss of air-to-ground communication is calculated by the following formulas (9) and (10);

[0140] The NLoS received power and the LoS received power are calculated by the following formulas (11) and (12).

[0141] Furthermore, the global optimal iteration unit 203 includes:

[0142] Initialization module is used to initialize the elite rate, determine the number of retained solutions and the number of updated solutions in each iteration according to the elite rate and the total number of preset solutions, and set the preset mutation probability p M ;

[0143] The loop control module is used to execute the following iterative loop until the maximum number of iterations is reached, or the iteration is stopped when the number of consecutive no-improvement times reaches the maximum number of consecutive no-improvement times:

[0144] The solution partitioning module is used to sort all the solutions, select the solutions with high fitness values ​​and the number of retained solutions as the solutions that remain unchanged among all the solutions, and select the solutions with low fitness values ​​and the number of updated solutions as the solutions that need to be updated among all the solutions;

[0145] A migration module, configured to perform a migration operation on each of the solutions that need to be updated, thereby updating the solutions that need to be updated;

[0146] A mutation module, configured to perform a mutation operation on the solution that needs to be updated after the migration operation is performed, thereby updating the solution that needs to be updated;

[0147] An evaluation and correction module, configured to evaluate and correct all solutions based on a connectivity constraint, wherein the connectivity constraint is configured to constrain each UAV to have at least two neighboring UAVs, where the neighboring UAVs are UAVs whose distance is less than a preset neighbor distance threshold;

[0148] a fitness value updating module configured to calculate, for each solution, the number of users that can be served by all the drones in the solution based on the position coordinates of all the drones in the solution and the preset user coordinates of multiple users, and to update the fitness value corresponding to the solution using the number of users that can be served by all the drones in the solution;

[0149] The optimal determination module is used to update the solution sorting by sorting all solutions in ascending order according to their corresponding updated fitness values, update the global optimal solution using the solution with the largest fitness value in the solution sorting, and update the global optimal fitness value using the fitness value corresponding to the global optimal solution; when stopping the iteration, output the global optimal solution and the global optimal fitness value.

[0150] Among them, for each solution, the number of users that can be served by all drones in the solution is calculated according to the position coordinates of all drones in the solution and the preset user coordinates of multiple users, and the fitness value corresponding to the solution is updated using the number of users that can be served by all drones in the solution. Reference can be made to the aforementioned embodiment in which, for each solution, the number of users that can be served by all drones in the solution is calculated according to the position coordinates of all drones in the solution and the preset user coordinates of multiple users, and the number of users that can be served by all drones in the solution is used as an explanation of the fitness value corresponding to the solution.

[0151] Furthermore, the migration module includes:

[0152] The migration rate determination module is used to set the migration rate and migration rate for all solutions according to the solution ranking of all solutions, wherein the solution with a larger fitness value corresponds to a larger migration rate; the migration rate and migration rate are both less than or equal to 1, and the sum of the migration rate and migration rate of the same solution is 1;

[0153] The migration operation module is used to generate a first random number corresponding to each solution in the interval (0, 1) for each solution that needs to be updated. If the first random number corresponding to the solution is less than the migration rate corresponding to the solution, the solution is used as the migration solution, and the horizontal coordinates of a drone randomly selected from the migration-out solution are replaced with the horizontal coordinates of the drone randomly selected from the migration-in solution; wherein the migration-out solution is selected from all solutions remaining after removing the migration-in solution using a roulette algorithm.

[0154] Furthermore, the migration rate determination module is configured as follows:

[0155] An ascending arithmetic sequence is generated between [0, 1], and terms in the arithmetic sequence are used as migration rates, wherein the number of terms in the arithmetic sequence is the same as the number of the multiple solutions;

[0156] According to the order of the solutions in the solution sorting and the terms in the arithmetic progression, the migration rate corresponding to each solution is obtained. The migration rate corresponding to each solution is equal to 1 minus the migration rate of the solution to one.

[0157] Furthermore, the mutation module is configured as follows:

[0158] For each solution that needs to be updated, generate a second random number from the interval (0, 1), and if the second random number is less than or equal to the preset mutation probability, use the solution as a mutation solution;

[0159] For each drone in the mutation solution, the position coordinates of the drone are replaced by position coordinates randomly generated in the candidate space.

[0160] Furthermore, the evaluation and correction module is configured as follows:

[0161] For each solution, each drone in the solution is used as a target drone, and the distance between the target drone and other drones in the solution is calculated. If the obtained distance is less than a preset neighbor distance threshold, the other drone is added as a neighbor drone of the target drone;

[0162] Check if each target drone has at least two neighbor drones;

[0163] If it is found that the target drone has fewer than two neighboring drones, the target drone is adjusted toward the nearest non-neighboring drone by a moving distance, with the line connecting the target drone as the direction, to satisfy the connectivity constraint. The moving distance is the actual distance between the target drone and the non-neighboring drone minus a preset neighbor distance threshold.

[0164] As for the device embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0165] The above technical solutions of the embodiments of the present invention are described in detail below with reference to specific application examples. For technical details not introduced during the implementation process, please refer to the relevant description above.

[0166] Problem Description

[0167] The dense urban landscape and irregular user distribution pose significant challenges to the construction of emergency communication networks, necessitating a stable and resilient emergency network deployment strategy. Drones, renowned for their agility, mobility, and cost-effectiveness, are increasingly being used in various military and civilian applications, including logistics and distribution, power inspection, reconnaissance and surveillance, and communications relay. The integration of drones with mobile base stations facilitates the rapid construction of wireless networks. Their rapid deployment, agile mobility, and high-quality connectivity offer an effective solution for communication restoration in post-disaster urban scenarios.

[0168] This embodiment of the present invention solves the problem of optimizing the coordinated deployment of multiple UAV base stations taking into account building obstacles.

[0169] When considering the optimization problem of multi-UAV (base station) coordinated deployment with building obstacles, such as Figure 3 As shown in the figure, buildings of varying heights are randomly distributed across the target area. Multiple drone base stations are deployed above the buildings, forming a dual-connectivity drone network that collaboratively provides communication services to users randomly distributed on the ground. If the communication path between the drone and the user is not blocked by buildings, the drone and the user communicate via a LoS link; otherwise, they communicate via an NLoS link.

[0170] The application scenarios of the embodiments of the present invention include the following main components:

[0171] Buildings: As one of the main obstacles, buildings are randomly distributed in the urban environment. They will not only occupy the potential deployment positions of drones in space and affect the solution space, but also block the transmission of wireless communication signals, affecting the quality of the user's received signal. The embodiment of the present invention describes a method for judging whether the communication path between a drone and a user is blocked by a building. Even users who are at the same distance from the drone base station will receive different signals due to differences in the distribution of buildings. The blocking effect of buildings can be handled through the coordinated cooperation of multiple drone base stations. If the communication link between the user and a drone base station is blocked by a building, other drone base stations whose communication links are not blocked can be selected to obtain network services. Therefore, compared with a single drone, multi-drone cooperation has a wider coverage range and larger capacity, which significantly improves service efficiency.

[0172] Drones: First, the dense distribution of buildings in urban environments occupies a large area of ​​feasible deployment space for drones. Second, drones provide communication services to ground users by carrying mobile base stations, and the communication model is generally established based on the air-to-ground channel model. Figure 4 As shown, due to building obstruction, the communication link between the drone and the user can be divided into LoS ​​links and NLoS links. Different communication links have different path losses, resulting in different communication quality received by the user. This embodiment of the present invention uses received power as an indicator to measure communication quality and describes the communication channel model and received power calculation method. In addition, due to broadband capacity limitations, the maximum number of users that each drone base station can serve is fixed. Finally, to provide stable and reliable communication services to ground users, drone base stations need to meet dual connectivity constraints to form a robust and flexible drone network.

[0173] Users: Users are randomly distributed on the two-dimensional surface of the target area and are assumed to be homogeneous, with the same minimum communication quality requirements. Communication service is guaranteed only when the user's received power reaches a preset threshold. In scenarios where multiple drones are operating collaboratively, users will select the drone with the strongest received power for communication access.

[0174] Obstruction judgment: When the drone base station communicates with users, if there are no obstacles blocking the transmission path, the signal propagates directly and the communication link is LoS. Conversely, when the transmission path is blocked by obstacles, the signal experiences reflection, refraction, and diffraction, and the link is NLoS. For a LoS link, users at the same distance from the drone base station experience the same signal attenuation and receive the same communication quality. However, building obstruction can affect the quality of the communication signal received by the users. If some users at the same distance from the drone base station have buildings blocking their communication path, that is, the communication link is NLoS, then these users will experience communication quality different from that of a LoS link.

[0175] In order to determine the type of communication link between the drone base station and the ground user, the four vertices and corresponding building heights in the building's top view can be used to represent the buildings in the target space. At the same time, the straddle method in computational geometry is used to determine whether there are obstacles between the drone base station and the ground user. The straddle method uses the mathematical properties of vector cross product and dot product to quickly determine whether two line segments intersect. The main principle of the method is as follows: Figure 5 As shown, there are two necessary and sufficient conditions for the intersection of line segments P1P2 and Q1Q2, namely formula (7) as condition 1 and formula (8) as condition 2. If these two conditions are met at the same time, line segment P1P2 intersects line segment Q1Q2.

[0176] Channel Model: To characterize the air-to-ground channel, a simple binary channel state model can be used, which only includes two link states: LoS link or NLoS link. If there are no buildings blocking the connection between the drone and the user, their communication link is considered to be a LoS link. Otherwise, the communication link between them is considered to be an NLoS link.

[0177] The power loss of air-to-ground communication under the binary channel model is expressed as formula (9) and formula (10);

[0178] In urban environments, buildings are densely distributed, and the additional path loss caused by the shadow effect of buildings must be considered. The additional random path loss under building shading conditions is calculated as s The embodiment of the present invention considers a relatively stable system, so the normal distribution random component is not used as the position variability. LoS and η NLoS The expression for η is LoS =92.4,η NLoS =92.4+L s , Where normrnd(μ,σ) represents the normal distribution, θ ij is the elevation angle of user i to drone j, and the rest are empirical parameters. c =2GHz, Table 1 shows the values ​​of empirical parameters under all environments.

[0179]

[0180] Table 1f c = Values ​​of empirical parameters under all conditions at 2GHz

[0181] After obtaining the path loss between the user and the UAV base station, the received power is equal to the transmitted power minus the path loss, and the calculation formula is formula (11) and formula (12);

[0182] For each user, the power received from different drone base stations needs to be calculated. Only when the power reaches a predetermined minimum threshold can the user receive communication service. The higher the power received by the user, the higher the communication quality.

[0183] Model:

[0184] This embodiment of the present invention solves a novel multi-UAV base station coordinated deployment optimization problem based on a binary channel model and MCLP model, taking into account deployment space restrictions, building obstructions, user communication conditions, UAV capacity limitations, and dual connectivity constraints. Based on this, the optimization problem is mathematically formalized as follows.

[0185] The meanings of the parameter variables used in the model are shown in Table 2:

[0186]

[0187]

[0188] Table 2 Question parameters

[0189] Objective function: The goal of the coordinated deployment optimization problem of multiple drone base stations is to maximize the number of users served. In order to maintain communication availability, the user's received power must exceed a given minimum threshold. The user prefers to select the drone base station that provides the highest received power for communication access. If the multiple drone base stations reach their service capacity, the user will not be able to access it and must select other drone base stations for communication access in descending order of received power. The embodiment of the present invention defines the decision variable b ik Indicates the user's access status:

[0190]

[0191] Therefore, the objective function can be expressed as

[0192] The constraints include: candidate location constraints, drone quantity constraints, feasibility constraints, uniqueness constraints, service capacity constraints, and connectivity constraints between drones.

[0193] ① Candidate position constraint: In an urban combat environment, the candidate position of the UAV is obtained by rasterizing the given space with a cube of 1 meter. Z represents the set of all grid positions in the target space. The candidate position k is determined by the center coordinate (x k ,y k ,z k ) represents the space. Due to the dense urban environment, drones can only be deployed in spaces outside buildings. The grid of buildings is B, and drones cannot be deployed within these grids. Considering the form of obstacles and the boundaries of the deployment space, the spatial coordinates of candidate location k are subject to the following constraints:

[0194]

[0195] ② UAV quantity constraint: This embodiment of the present invention primarily establishes a multi-UAV base station coordinated deployment model based on a maximum coverage location model. Therefore, it is necessary to fully utilize the given UAV resources. The final number of deployed UAVs should be limited to n.

[0196]

[0197] ③ Feasibility constraint: For each user, when a drone provides communication services to it, its receiving power must reach a certain threshold. Therefore, if it is determined that drone k serves user i, then user i gets the receiving power P of drone k ik Must be greater than or equal to the minimum received power threshold P min The key constraint can be described as:

[0198]

[0199] ④ Uniqueness constraint: Each user can only access one drone base station at most, that is:

[0200]

[0201] ⑤ Service capacity constraint: For each drone base station, the number of users it can access cannot exceed its service capacity, that is:

[0202]

[0203] ⑥ Connectivity constraints between drones

[0204] For each drone, it needs to be able to find at least one max Two other drones. Let the two drones be i and j, and the distance between them is d ij . Use 0-1 variable c ij Indicates the connectivity between drones:

[0205]

[0206] The connectivity constraint between UAVs is expressed as:

[0207] Improved biogeography optimization algorithm: Based on the characteristics of intelligent drone clusters, an embodiment of the present invention proposes an improved biogeography optimization (BBO) algorithm for solving the optimization problem of coordinated deployment of multiple drone base stations. The BBO algorithm is inspired by biogeography theory, which simulates the species migration process between multiple habitats. The suitability of each habitat is represented by the habitat suitability index (HSI) and is affected by suitability index variables (SIVs) such as temperature, rainfall, and soil quality. Habitats with high HSI allow more species to exist, while habitats with low HSI have fewer species, and HSI is determined by SIV. In the implementation of the BBO algorithm, two main operators, migration and mutation, are included, which are very suitable for solving combinatorial optimization problems. The pseudo code of the improved BBO algorithm is detailed in Algorithm 1.

[0208]

[0209]

[0210] Solution Representation: To address the multi-UAV base station deployment problem, we redefined the solution representation. Each solution contains the spatial Cartesian coordinates of all UAV base stations, and the solution's fitness is the number of ground users served. Based on the concepts of the BBO algorithm, each solution is considered a habitat, where the spatial Cartesian coordinates of each UAV are SIV, and the solution's fitness is HIS.

[0211] The fitness level of a solution reflects the size of the population. Sorting all solutions from lowest to highest fitness level, the solution with the lowest fitness represents a habitat with a zero population size, an in-migration rate of 1, and a zero out-migration rate. Conversely, the solution with the highest fitness represents a habitat with the largest population size, a zero in-migration rate, and a one-migration rate. For the other solutions, the population size of the habitats increases as fitness increases, while the in-migration rate decreases and the out-migration rate increases. In practical applications of this algorithm, the changes in the in-migration rate and the out-migration rate can be set to be linear, with the sum of the two being 1.

[0212] Initial solution construction: The deployment of multiple UAV base stations is restricted by the boundaries of the target space and the building space. Therefore, the candidate locations must be selected from the barrier-free space, and when generating a solution, the positions of the UAVs in the solution must be adjusted to satisfy the dual connection constraints between the UAVs. The embodiment of the present invention proposes two strategies for constructing the initial solution. The first method is to randomly generate the three-dimensional coordinates of multiple UAVs in the available deployment space. The second method uses the k-means clustering method to generate the horizontal coordinates of multiple UAVs in the deployable space, and then randomly determines the height (H) within the specified range. Kmin, H Kmax ). The BBO algorithm based on random initialization is denoted as RD+BBO, and the BBO algorithm based on k-means initialization is denoted as KMC+BBO.

[0213] Elite strategy: The main idea of ​​the BBO algorithm is to use good solutions to enhance poor solutions. Elite strategy is a commonly used technique that ensures that excellent solutions can be retained to the next generation. Elite solutions refer to the solutions that perform best in the current population. These solutions are usually retained directly to the next generation to maintain or improve the solution quality of the entire population. For this purpose, the retention rate is introduced, represented by β∈[0,1), which determines the proportion of elite solutions retained in each iteration. The lower β means the fewer solutions are retained. Assume that the total number of solutions is s. The number of elite solutions retained is s Keep =βs. New solutions can be improved through migration and mutation to obtain better solutions. The calculation method of the number of new solutions can be expressed as sNew =ss Keep =(1-β)s. The purpose of the elite strategy is to maintain the diversity of the population while retaining excellent individual information to prevent the algorithm from converging to a local optimal solution too early.

[0214] Migration operation: The migration operator is a core operator in the BBO algorithm and is crucial for constructing new solutions. It mainly transfers the characteristics of the solution between habitats through the probability defined by the immigration rate λ(i) and the emigration rate μ(i). The BBO algorithm uses the immigration rate and emigration rate as linear functions of species counts, as shown in the following formula:

[0215]

[0216] Where I and E are the maximum immigration rate and emigration rate, respectively, which are usually set to 1. The symbol S represents the number of species, and S max Indicates the maximum number of species.

[0217] The immigration rate λ(i) is used to select a solution x i , the migration rate μ(j) is used to select another solution x j Generate a random number τ in the interval (0,1). If τ is less than the migration rate λ(i), perform the migration operation on solution i. This operation includes i Select two SIVs (horizontal coordinates of the drone position) and use them to solve x j Select the corresponding SIV to replace it.

[0218] Mutation operator: The mutation operator in the BBO algorithm simulates the changes in the habitat environment. The mutation probability of a habitat is a function of the number of species in the habitat, expressed as follows:

[0219]

[0220] Where m s is the mutation probability when the number of species is s, m max is the maximum mutation rate, P s is the probability that the number of species is s, P max P s The maximum value of .

[0221] In the BBO algorithm, the original mutation operator involves a complex determination of the mutation probability of each solution. To simplify this, a constant mutation probability p can be applied to the coordinates of each drone. M ∈(0,1). p M The value of should be close to zero, indicating that the drone position is less sensitive to mutations. Generate a random number σ from the interval (0,1). If σ≤p M, then perform mutation operation on the corresponding solution. Mutation operation refers to generating a new drone coordinate in the available target space to replace the original drone coordinate.

[0222] Improved Biogeographic Optimization Algorithm: To further analyze the effectiveness of the proposed method and its feasibility in real-world applications, the Beichen Delta region of Changsha was selected as a case study for testing. The main experimental parameters are shown in Table 3.

[0223] Case description: The Beichen Delta region is located in Changsha, China, covering an area of ​​1.1 square kilometers. Figure 6 The area outlined by the black line represents a bird's-eye view of the Beichen Delta. We extracted buildings over 50 meters and selected 35 of them with relatively regular shapes. We also simulated 60 users with communication needs, randomly distributed within the target area.

[0224] The distribution of buildings and users is as follows Figure 7 shown.

[0225]

[0226] Table 3 Experimental parameters

[0227] Results and Discussion: In this example, six drone base stations are deployed to provide wireless network services to 60 users. Both the BBO algorithm with random initialization (RD+BBO) and the BBO algorithm with k-means initialization (KMC+BBO) are used to solve this example. The maximum number of users covered by both methods is 60.

[0228] The optimal deployment results of RD+BBO method in three dimensions and two dimensions are as follows: Figure 8 and Figure 9 As shown. The optimal results of the KMC+BBO method in three dimensions and two dimensions are as follows Figure 10 and Figure 11 The convergence curves of the two methods are shown in Figure 12 As shown in the figure, the experimental results and convergence curves show that the proposed method can achieve effective coordinated deployment of multiple UAV base stations with architectural obstacles. The KMC+BBO method not only produces optimal results but also exhibits faster convergence, which is attributed to the high quality of its initial solution. Therefore, the KMC+BBO method is more suitable for solving the coordinated deployment optimization problem of multiple UAV base stations.

[0229] It should be understood that the specific order or hierarchy of steps in the disclosed processes is an example of an exemplary method. Based on design preferences, it should be understood that the specific order or hierarchy of steps in the process can be rearranged without departing from the scope of the present disclosure. The accompanying method claims present elements of the various steps in an exemplary order and are not intended to be limited to the specific order or hierarchy described.

[0230] In the foregoing detailed description, various features are grouped together in a single embodiment to simplify the disclosure. This method of disclosure should not be interpreted as reflecting an intention that embodiments of the claimed subject matter require more features than are expressly recited in each claim. On the contrary, as reflected in the appended claims, the invention comprises less than all the features of any individual disclosed embodiment. The appended claims are hereby expressly incorporated into the detailed description, with each claim standing on its own as a separate preferred embodiment of the invention.

[0231] The above description of the disclosed embodiments is intended to enable any person skilled in the art to implement or use the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be applied to other embodiments without departing from the spirit and scope of the present disclosure. Therefore, the present disclosure is not limited to the embodiments presented herein but is intended to be accorded the widest scope consistent with the principles and novel features disclosed herein.

[0232] The above description includes examples of one or more embodiments. Of course, it is impossible to describe all possible combinations of components or methods for the purpose of describing the above embodiments, but it will be appreciated by those skilled in the art that the various embodiments may be further combined and arranged. Therefore, the embodiments described herein are intended to encompass all such changes, modifications and variations that fall within the scope of protection of the appended claims. In addition, to the extent the term "comprising" is used in the specification or claims, the term is intended to be encompassed in a manner similar to the term "including". In addition, any use of the term "or" in the specification of the claims is intended to mean a "non-exclusive or".

[0233] Those skilled in the art will also appreciate that the various illustrative logical blocks, units, and steps listed in the embodiments of the present invention can be implemented by electronic hardware, computer software, or a combination of the two. To clearly demonstrate the interchangeability of hardware and software, the various illustrative components, units, and steps described above have generally described their functions. Whether such functions are implemented by hardware or software depends on the specific application and the design requirements of the entire system. Those skilled in the art may use various methods to implement the described functions for each specific application, but such implementation should not be understood as exceeding the scope of protection of the embodiments of the present invention.

[0234] The various illustrative logic blocks or units described in the embodiments of the present invention can be implemented or operated by a general-purpose processor, a digital signal processor, an application-specific integrated circuit (ASIC), a field programmable gate array or other programmable logic device, discrete gate or transistor logic, discrete hardware components, or any combination thereof. The general-purpose processor can be a microprocessor, and optionally, the general-purpose processor can also be any conventional processor, controller, microcontroller or state machine. The processor can also be implemented by a combination of computing devices, such as a digital signal processor and a microprocessor, a plurality of microprocessors, one or more microprocessors combined with a digital signal processor core, or any other similar configuration.

[0235] The steps of the methods or algorithms described in the embodiments of the present invention may be directly embedded in hardware, a software module executed by a processor, or a combination of the two. The software module may be stored in RAM memory, flash memory, ROM memory, EPROM memory, EEPROM memory, registers, a hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art. For example, the storage medium may be connected to the processor so that the processor can read information from the storage medium and write information to the storage medium. Alternatively, the storage medium may also be integrated into the processor. The processor and storage medium may be provided in an ASIC, which may be provided in a user terminal. Alternatively, the processor and storage medium may also be provided in different components in the user terminal.

[0236] In one or more exemplary designs, the above-mentioned functions described in the embodiments of the present invention can be implemented in hardware, software, firmware, or any combination of the three. If implemented in software, these functions can be stored on a computer-readable medium or transmitted in the form of one or more instructions or codes on a computer-readable medium. Computer-readable media include computer storage media and communication media that facilitate the transfer of computer programs from one location to another. Storage media can be any available medium that can be accessed by a general or special computer. For example, such computer-readable media can include but are not limited to RAM, ROM, EEPROM, CD-ROM or other optical disk storage, magnetic disk storage or other magnetic storage devices, or any other medium that can be used to carry or store program code in the form of instructions or data structures and other forms that can be read by a general or special computer, or a general or special processor. In addition, any connection can be appropriately defined as a computer-readable medium. For example, if the software is transmitted from a website, server or other remote resource via a coaxial cable, fiber optic cable, twisted pair, digital subscriber line (DSL), or wireless methods such as infrared, wireless, and microwave, it is also included in the definition of computer-readable media. The disks and discs mentioned above include compact disks, laser disks, optical disks, DVDs, floppy disks, and Blu-ray discs. Disks typically reproduce data magnetically, while discs typically reproduce data optically with lasers. Combinations of the above may also be included in computer-readable media.

[0237] The specific implementation methods described above further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific implementation method of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for optimizing the coordinated deployment of multiple UAV base stations under building obstacles, characterized in that: include: Randomly assign position coordinates to all drones in the candidate space multiple times to obtain multiple solutions, each of which includes the position coordinates of all drones; The candidate space is a space not occupied by buildings in the environment where the drone is deployed; For each solution, the number of users that can be served by all the drones in the solution is calculated based on the position coordinates of all the drones in the solution and the preset user coordinates of multiple users, and the number of users that can be served by all the drones in the solution is used as the fitness value corresponding to the solution; All solutions are sorted in ascending order according to their corresponding fitness values ​​to obtain a solution ranking, the solution with the largest fitness value in the solution ranking is taken as the global optimal solution, and the fitness value corresponding to the global optimal solution is taken as the global optimal fitness value; Based on the improved biogeographic optimization algorithm, the multiple solutions are iteratively updated, and the multiple solutions are sorted according to the fitness values ​​of the multiple solutions updated in each iteration. The global optimal solution and the global optimal fitness value are updated according to the obtained sorting. The iteration is stopped until the number of cyclic iterations reaches a preset maximum number of iterations or the number of consecutive no-improvement times reaches a preset maximum number of consecutive no-improvement times. The global optimal solution is used for the coordinated deployment of multiple UAV base stations. The obtained global optimal fitness value is the number of users that can be served after the coordinated deployment of multiple UAV base stations according to the obtained global optimal solution; The improved biogeographic optimization algorithm is used to iteratively update the multiple solutions, including: evaluating and correcting all solutions based on connectivity constraints, where the connectivity constraints are used to constrain each drone to have at least two neighboring drones, where the neighboring drones are drones whose distance is less than a preset neighbor distance threshold; applying a constant mutation probability pM∈(0,1) to the coordinates of each drone, generating a random number σ from the interval (0,1), and performing a mutation operation on the corresponding solution if σ≤pM; the mutation operation refers to generating a new drone coordinate in the target space to replace the original drone coordinate.

2. The method for optimizing the coordinated deployment of multiple UAV base stations under architectural obstacles according to claim 1, wherein: The multiple random assignments of position coordinates for all drones in the candidate space are performed to obtain multiple solutions, each solution including the position coordinates of all drones, including: Acquiring spatial information of an environmental space, rasterizing the environmental space into a plurality of continuously arranged cubic grids, and obtaining a grid set of buildings within the environmental space and a grid set of candidate spaces; the grid set comprising a plurality of cubic grids; the spatial information of the environmental space comprising: the length, width, and height of the environmental space, and spatial information of one or more buildings within the environmental space; the spatial information of the buildings comprising the location coordinates, length, width, and height of the buildings; Generating the three-dimensional position coordinates of all drones in each solution in a random manner within the grid set of the candidate space; and / or, The k-means clustering method is used to generate the horizontal coordinates of all drones in the candidate space, and then the height coordinates of all drones are randomly determined within the preset candidate space height range; the position coordinates of a drone are the center coordinates of the cubic grid occupied by the drone; the position coordinates include horizontal coordinates and height coordinates.

3. The method for optimizing the coordinated deployment of multiple UAV base stations under architectural obstacles according to claim 2, wherein: The k-means clustering method is used to generate the horizontal coordinates of all drones in the candidate space, and then the height coordinates of all drones are randomly determined within the preset candidate space height range, including: Performing k-means clustering on the multiple users according to horizontal coordinates in the preset user coordinates to obtain horizontal coordinates of multiple cluster centers, and using the horizontal coordinates of the multiple cluster centers as the horizontal coordinates of the multiple drones; For each drone, based on a probabilistic statistical channel model, determine the optimal deployment altitude corresponding to the drone's maximum coverage radius. With the optimal deployment altitude as the center, determine the upper and lower limits of a preset candidate spatial altitude range for the drone's altitude coordinate, where the upper limit is equal to the optimal deployment altitude plus a preset upward offset value, and the lower limit is equal to the optimal deployment altitude minus a preset downward offset value. The probabilistic statistical channel model is a model expressed by the following formula: Where h is the altitude of the drone, d ik is the distance from user i to the center of the coverage circle of drone k, a and b are constant coefficients determined by the environment, is the elevation angle from UAV k to user i, ηLoS and ηnLoS are the average additional losses of LoS and nLoS respectively, f c is the carrier frequency of the air-to-ground channel and c is the speed of light.

4. The method for optimizing the coordinated deployment of multiple UAV base stations under architectural obstacles according to claim 1, wherein: For each solution, the number of users that can be served by all the drones in the solution is calculated based on the position coordinates of all the drones in the solution and the preset user coordinates of multiple users, and the number of users that can be served by all the drones in the solution is used as the fitness value corresponding to the solution, including: For each solution, calculate whether there is a building obstruction between each user and each drone in the solution based on the preset user coordinates of each user, the location coordinates of each drone in the solution, and the spatial information of all buildings in the environment space; the spatial information of the building includes the location coordinates, length, width, and height of the building; For each user, for links between drones with building obstructions, calculate the NLoS received power (NROP) obtained by the user from the drones with building obstructions. For links between drones with no building obstructions, calculate the LoS received power obtained by the user from the drones without building obstructions. For each user, the user is assigned to a drone corresponding to the maximum value of all the NLoS reception powers and LoS reception powers of the user; For each solution, the total number of users served by all drones in the solution is counted as the fitness value corresponding to the solution.

5. The method for optimizing the coordinated deployment of multiple UAV base stations under architectural obstacles according to claim 4, wherein: The calculation of the NLoS received power obtained by the user from the drone with building obstruction for the link between the user and the drone with building obstruction, and the calculation of the LoS received power obtained by the user from the drone without building obstruction for the link between the user and the drone with building obstruction, specifically includes: Based on the binary channel model, the power loss of air-to-ground communication is calculated by the following formula: EN LoS =20lg(d ij )+20lg(f c )+η LoS EN NLoS =20lg(d ij )+20lg(f c )+η NLoS Among them, f c is the carrier frequency, d ij is the distance between user i and drone j, η LoS and η NLoS represents the additional path power loss under link state LoS and NLoS respectively; η LoS =92.4,η NLoS =92.4+L s , additional random path loss under building occlusion conditions normrnd(μ,σ) represents normal distribution, θ ij is the elevation angle of user i to drone j, g μ 、g σ 、h μ 、h σ 、i μ 、i σ is an empirical parameter; The NLoS received power and LoS received power are calculated using the following formula: P LoS =P t -PL LoS P NLoS =P t -PL NLoS Among them, P t Transmit power to the drone, PL LoS and PL NLoS represent the path loss under LoS and NLoS links respectively, P LoS and P NLoS They represent the user's LoS received power and NLoS received power when the link status is LoS and NLoS, respectively.

6. The method for optimizing the coordinated deployment of multiple UAV base stations under architectural obstacles according to claim 1, wherein: The improved biogeographic optimization algorithm is based on loop iteration and updating of the multiple solutions. The multiple solutions are sorted according to the fitness values ​​of the multiple solutions updated in each iteration, and the global optimal solution and the global optimal fitness value are updated according to the obtained sorting. The iteration is stopped until the number of loop iterations reaches a preset maximum number of iterations or the number of consecutive no-improvement times reaches a preset maximum number of consecutive no-improvement times. The obtained global optimal solution is used for the coordinated deployment of multiple UAV base stations. The obtained global optimal fitness value is the number of users that can be served after the coordinated deployment of multiple UAV base stations according to the obtained global optimal solution, including: Initialize the elite rate, determine the number of retained solutions and the number of updated solutions for each iteration based on the elite rate and the total number of preset solutions, and set the preset mutation probability p M ; The following iterative loop is executed until the maximum number of iterations is reached, or the iteration is stopped when the number of consecutive no-improvement times reaches the maximum number of consecutive no-improvement times: According to the sorting of all solutions, the solutions with high fitness values ​​and the number of retained solutions are taken as the solutions that remain unchanged among all solutions, and the remaining solutions with low fitness values ​​and the number of updated solutions are taken as the solutions that need to be updated among all solutions; Performing a migration operation on each of the solutions that need to be updated, and updating the solutions that need to be updated; Performing a mutation operation on the solution that needs to be updated after the migration operation is performed, so as to update the solution that needs to be updated; All solutions are evaluated and modified based on connectivity constraints, where each UAV is constrained to have at least two neighboring UAVs, where the neighboring UAVs are UAVs whose distance is less than a preset neighbor distance threshold. For each solution, the number of users that can be served by all the drones in the solution is calculated based on the position coordinates of all the drones in the solution and the preset user coordinates of multiple users, and the fitness value corresponding to the solution is updated using the number of users that can be served by all the drones in the solution; Sorting all solutions in ascending order of their corresponding updated fitness values, updating the solution sorting, using the solution with the largest fitness value in the solution sorting to update the global optimal solution, and using the fitness value corresponding to the global optimal solution to update the global optimal fitness value; When the iteration is stopped, the global optimal solution and the global optimal fitness value are output.

7. The method for optimizing the coordinated deployment of multiple UAV base stations under architectural obstacles according to claim 6, wherein: Performing a migration operation on each of the solutions that need to be updated to update the solutions that need to be updated includes: According to the solution ranking of all solutions, set the immigration rate and emigration rate for all solutions, among which the solution with a large fitness value corresponds to a large immigration rate; the immigration rate and emigration rate are both less than or equal to 1, and the sum of the immigration rate and emigration rate of the same solution is 1; For each solution that needs to be updated, a first random number corresponding to the solution is generated in the interval (0, 1). If the first random number corresponding to the solution is less than the immigration rate corresponding to the solution, the solution is used as the immigration solution, and the horizontal coordinates of a drone randomly selected from the outgoing solution are replaced with the horizontal coordinates of the drone randomly selected from the incoming solution. The outgoing solution is selected using a roulette algorithm from all solutions remaining after removing the incoming solution.

8. The method for optimizing the coordinated deployment of multiple UAV base stations under architectural obstacles according to claim 7, wherein: The step of setting the migration rate and the migration rate for all solutions according to the solution sorting includes: An ascending arithmetic sequence is generated between [0, 1], and terms in the arithmetic sequence are used as migration rates, wherein the number of terms in the arithmetic sequence is the same as the number of the multiple solutions; According to the order of the solutions in the solution sorting and the terms in the arithmetic progression, the migration rate corresponding to each solution is obtained. The migration rate corresponding to each solution is equal to 1 minus the migration rate of the solution to one.

9. The method for optimizing the coordinated deployment of multiple UAV base stations under architectural obstacles according to claim 6, wherein: All solutions are evaluated and corrected based on connectivity constraints, including: For each solution, each drone in the solution is used as a target drone, and the distance between the target drone and other drones in the solution is calculated. If the obtained distance is less than a preset neighbor distance threshold, the other drone is added as a neighbor drone of the target drone; Check if each target drone has at least two neighbor drones; If it is found that the target drone has fewer than two neighboring drones, the target drone is adjusted toward the nearest non-neighboring drone by a moving distance, with the line connecting the target drone as the direction, to satisfy the connectivity constraint. The moving distance is the actual distance between the target drone and the non-neighboring drone minus a preset neighbor distance threshold.

10. A multi-UAV base station coordinated deployment optimization device under building obstacle conditions, characterized in that: include: an initial solution determination unit, configured to randomly assign position coordinates to all drones in a candidate space multiple times to obtain multiple solutions, each solution including the position coordinates of all drones; the candidate space being a space in the environment where the drones are deployed that is not occupied by buildings; a fitness value determining unit, configured to calculate, for each solution, the number of users that can be served by all the drones in the solution based on the position coordinates of all the drones in the solution and the preset user coordinates of a plurality of users, and use the number of users that can be served by all the drones in the solution as the fitness value corresponding to the solution; a global optimal determination unit, configured to sort all solutions in ascending order of their corresponding fitness values ​​to obtain a solution ranking, taking the solution with the largest fitness value in the solution ranking as the global optimal solution, and taking the fitness value corresponding to the global optimal solution as the global optimal fitness value; A global optimal iteration unit is configured to perform cyclic iterative updates on the multiple solutions based on an improved biogeographic optimization algorithm, and to sort the multiple solutions according to the fitness values ​​of the multiple solutions updated in each iteration, and to update the global optimal solution and the global optimal fitness value according to the obtained sorting, until the number of cyclic iterations reaches a preset maximum number of iterations or the number of consecutive no-improvement times reaches a preset maximum number of consecutive no-improvement times, and to stop the iteration. The obtained global optimal solution is used for the coordinated deployment of multiple UAV base stations, and the obtained global optimal fitness value is the number of users that can be served after the coordinated deployment of multiple UAV base stations according to the obtained global optimal solution; The improved biogeographic optimization algorithm is used to iteratively update the multiple solutions, including: evaluating and correcting all solutions based on connectivity constraints, where the connectivity constraints are used to constrain each drone to have at least two neighboring drones, where the neighboring drones are drones whose distance is less than a preset neighbor distance threshold; applying a constant mutation probability pM∈(0,1) to the coordinates of each drone, generating a random number σ from the interval (0,1), and performing a mutation operation on the corresponding solution if σ≤pM; the mutation operation refers to generating a new drone coordinate in the target space to replace the original drone coordinate.

Images