An unmanned aerial vehicle positioning and deployment method based on scale-free properties for task scheduling

Abstract

This invention discloses a drone positioning and deployment method based on scale-free attribute-oriented task scheduling. This method is applied in scenarios involving multiple drones and multiple users, where the scenario includes numerous drones and users, and the drones provide edge computing services to the users. The method constructs a scale-free distance-weighted network, determines the central node of the network, and identifies the non-central nodes associated with each central node. Each module determines the position of each drone, and Weiszfeld is used for iterative updates, ultimately achieving the final positioning and deployment of each drone. When calculating the score for each node to become a central node, this method considers both node centrality and specificity, ensuring that users within the same class are close while users between classes are far apart, resulting in high accuracy.

Description

Technical Field

[0001] This invention belongs to the field of unmanned aerial vehicle (UAV) deployment technology, specifically relating to a UAV positioning and deployment method based on scale-free attribute-oriented task scheduling. Background Technology

[0002] In recent years, with the continuous development of Internet of Things (IoT) technology and the expansion of application scenarios, the computing tasks required by smart devices have gradually increased, exhibiting a clear demand for real-time processing. To address the battery limitations and computing pressure on IoT devices, mobile edge computing (MEC) has emerged as a solution for performing computing and processing tasks at the edge layer, closer to the terminal. Traditional mobile edge computing technology places servers near IoT devices, providing computing services and significantly improving the computing efficiency of IoT devices, thus promoting the development of IoT technology.

[0003] Unmanned Aerial Vehicle (UAV)-assisted mobile edge computing is increasingly becoming a new trend in industry development. By introducing UAVs with computing, storage, and communication capabilities as mobile edge nodes into edge networks, large-scale data from IoT devices can be dynamically covered and processed. Compared to traditional fixed-deployment edge servers, UAVs can flexibly adjust the geographical location and coverage of edge computing resources according to task load, regional coverage requirements, and network conditions, reducing transmission latency and bandwidth pressure. Through air-ground collaboration, UAVs can compensate for the insufficient coverage of fixed base stations, forming a multi-layered, heterogeneous computing network with ground edge servers. Edge tasks can be initially processed on UAVs, with key data filtered before transmission, reducing overall energy consumption and network transmission costs, and improving the system's fault tolerance to single points of failure.

[0004] Compared to traditional mobile edge computing, drone-assisted mobile edge computing faces more complex challenges. This is because drones themselves have limited computing power and communication coverage, and the mobility of users brings dynamism and variability. To achieve optimal resource utilization, a comprehensive strategy is needed to integrate drone trajectory control, strategic task offloading, and computing resource optimization. These strategies are interdependent and work together to form a collaborative optimization system, aiming to maximize resource utilization, comprehensively improve system performance, and robustly adapt to dynamic environments.

[0005] Existing drone positioning and deployment methods include heuristic algorithms and some common clustering algorithms. These algorithms can locate drones based on the current user distribution, but they have certain limitations. Heuristic algorithms, such as the frog-leapfrog algorithm, require randomly selecting a certain number of nodes as initial points, and subsequent steps have a high degree of randomness, resulting in poor algorithm stability. In drone deployment scenarios, due to dynamic changes in the environment, such as the increase or decrease in the number of drones or users, heuristic algorithms and common clustering algorithms, such as the K-Means algorithm, need to recalculate based on the new environment, making rapid deployment impossible. Summary of the Invention

[0006] The purpose of this invention is to address the problems raised in the background art by proposing a UAV positioning and deployment method based on scale-free attribute-oriented task scheduling.

[0007] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0008] This invention proposes a UAV localization and deployment method based on scale-free attribute-oriented task scheduling. This method is applied in scenarios involving multiple UAVs and multiple users.

[0009] The scene contains A drone and Individual users, and drones provide edge computing services to users;

[0010] Will A scale-free distance-weighted network is constructed for each user, with each user as a node in the scale-free distance-weighted network. The edges connecting any two nodes are pruned, and the remaining edges form the edges of the scale-free distance-weighted network. The weight of each edge is then calculated.

[0011] Determine the scale-free distance-weighted network One central node;

[0012] Determine the central node associated with each non-central node, and then... Each node is divided into Each module contains a central node and non-central nodes associated with that central node, utilizing... Module calculation The location of each drone is determined and updated iteratively until the coordinates of each drone remain stable. The final stable coordinates of each drone are then used as the final deployment location of the drone.

[0013] Preferably, within the same time slot, the location of each user is fixed, while between different time slots, each user moves. .

[0014] Preferably, when the Euclidean distance between the two nodes corresponding to the connecting edge is greater than a preset value, the current connecting edge is pruned.

[0015] Preferably, the formula for calculating the edge weights of the scale-free distance-weighted network is as follows:

[0016] ;

[0017] in, For the first The node and the first The weight of the edge formed by the nodes. Let be the set of all nodes in a scale-free distance-weighted network. For the first The node and the first Euclidean distance between nodes The maximum distance at which drones can provide edge computing services to users.

[0018] Preferably, the determination of the scale-free distance-weighted network The central nodes include:

[0019] Normalize the amount of tasks generated by each node;

[0020] The centrality of each node is obtained by weighting and summing the degree of each node with the normalized task amount of the corresponding node. The degree of each node is the sum of the edge weights formed by all other nodes in the scale-free distance weighted network and the current node.

[0021] Calculate the score by which each node becomes the central node based on the specific characteristics of each node;

[0022] Select the node with the highest centrality among all nodes. Centrality, and the former The nodes corresponding to the centrality are taken as the central nodes, and then we obtain One central node.

[0023] Preferably, the formula for calculating the fraction of each node becoming the center node is:

[0024] ;

[0025] in,

[0026] ;

[0027] in, For the first The score of each node becoming the central node For the first Node specificity For the first The node and the first Jaccard similarity of nodes, For the first Centrality of each node The set consisting of the centrality of all nodes. This indicates taking the maximum value. This indicates taking the minimum value.

[0028] Preferably, for each non-central node, the central node with the highest similarity to the current non-central node Jaccard is selected as the central node associated with the current non-central node.

[0029] Preferably, Each module and Each drone corresponds to a specific drone;

[0030] For each module, the coordinates of all nodes contained in each module are averaged to obtain the initial position of each UAV;

[0031] The positions of each drone are updated iteratively until the Euclidean distance between the positions of two consecutive updates is less than a set threshold. If the update iteration stops, the position of each drone corresponding to the last update iteration will be taken as the final deployment location of the drone, and the formula for updating the position of each drone is:

[0032] ;

[0033] in, For the first The next update iteration changes the current drone's x-coordinate or y-coordinate. For the first In the next update iteration, the module corresponding to the current drone is the [number]th [item]. The x-coordinate or y-coordinate of each node, This represents the total number of nodes in the module corresponding to the current drone. For the first In the next update iteration, the module corresponding to the current drone is the [number]th [item]. The distance between each node and the current drone. For the first The next update iteration includes all nodes in the module corresponding to the current drone and the... The reciprocal of the sum of the distances of the current drones during the next update iteration.

[0034] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0035] This scale-free attribute-based task scheduling-oriented UAV positioning and deployment method constructs a scale-free distance-weighted network, determines the central node of the scale-free distance-weighted network and the non-central nodes associated with each central node, uses each module to determine the position of each UAV, and uses Weiszfeld to iteratively update, thereby achieving the final positioning and deployment of each UAV.

[0036] This method considers both node centrality and specificity when calculating the score for each node to become a central node, ensuring that users within the same class are close while users between classes are far apart, resulting in high accuracy. Furthermore, the calculation of node centrality takes into account both the distance between users and the amount of tasks generated by each user. When the number of drones is limited, priority is given to users who generate a large amount of tasks, thereby improving the subsequent data calculation workload of the entire system. Attached Figure Description

[0037] Figure 1 This is a flowchart illustrating the UAV positioning and deployment method based on scale-free attribute-oriented task scheduling according to the present invention.

[0038] Figure 2 This is a schematic diagram of a deployment scenario for the UAV positioning and deployment method based on scale-free attribute-oriented task scheduling according to the present invention.

[0039] Figure 3 This is a schematic diagram illustrating the positioning and deployment of the UAV according to the present invention;

[0040] Figure 4 This is a schematic diagram comparing the effects of the method of the present invention with two existing methods. Detailed Implementation

[0041] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0042] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used herein in the description of the invention is for the purpose of describing particular embodiments only and is not intended to limit the invention.

[0043] like Figures 1-4 As shown, a scale-free attribute-based task scheduling-oriented UAV positioning and deployment method is provided for application in scenarios with multiple UAVs and multiple users, wherein:

[0044] Step 1, the scene contains A drone and There are 20 users (e.g., 3 drones, 20 users), and the drones provide edge computing services to the users; within the same time slot, the location of each user is fixed, but between different time slots, the users move. The drone provides edge computing services to users, receiving data unloaded by the user, performing data calculations, and returning the results. The drone's signal coverage is limited, and it can only provide services to users within its coverage area. The maximum horizontal distance of the drone's service range is [missing information]. The altitude of each drone remains fixed and all drones maintain the same altitude, such as 60 meters. Within each time slot, the drones perform addressing and positioning based on the user's location distribution and the amount of tasks generated by each user, and provide computing services.

[0045] Step 2, A scale-free distance-weighted network is constructed for each user, with each user as a node in the scale-free distance-weighted network. The edges connecting any two nodes are pruned, and the remaining edges form the edges of the scale-free distance-weighted network. The weight of each edge is then calculated.

[0046] Specifically, when the Euclidean distance between the two nodes corresponding to the connecting edge is greater than a preset value (in this embodiment, the preset value is...), the Euclidean distance between the two nodes is greater than a preset value. If the current edge of the connection is pruned, then the current edge will be pruned.

[0047] The formula for calculating the edge weights of the scale-free distance-weighted network is as follows:

[0048] ;

[0049] in, For the first The node and the first The weight of the edge formed by the nodes. Let be the set of all nodes in a scale-free distance-weighted network. For the first The node and the first Euclidean distance between nodes ( , For the first The coordinates of each node, For the first (coordinates of each node) Given the maximum distance (e.g., 150 meters) at which a drone can provide edge computing services to a user, the edge set of the scale-free distance-weighted network is: .

[0050] Step 3: Determine the scale-free distance-weighted network. The central nodes include:

[0051] Step 3.1: Normalize the amount of tasks generated by each node (this normalization is min-max normalization, the formula is: ,in, For the scale-free distance-weighted network, the first The normalized task quantity for each node , , In order, the scale-free distance-weighted networks are as follows: The amount of tasks generated by each node For the scale-free distance-weighted network, the first The amount of tasks generated by each node.

[0052] Step 3.2: Calculate the centrality of each node by weighted summing of its degree and the normalized task amount. The degree of each node is the sum of the edge weights formed by all other nodes in the scale-free distance-weighted network and the current node. The formula for calculating the centrality of each node is as follows:

[0053] ;

[0054] in,

[0055] ;

[0056] in, For the first The degree of each node, Weights (set based on experience).

[0057] Step 3.3: Calculate the score for each node to become a central node based on the specificity of each node. The formula for calculating the score for each node to become a central node is as follows:

[0058] ;

[0059] in,

[0060] ;

[0061] ;

[0062] in, For the first The score of each node becoming the central node For the first Node specificity For the first The node and the first Jaccard similarity of nodes, For the first Centrality of each node The set consisting of the centrality of all nodes. This indicates taking the maximum value. This indicates taking the minimum value. For the first All neighboring nodes of a node, For the first All neighboring nodes of a node, , For the first All neighboring nodes of the i-th node and the i-th node The first node in the common set of all its neighboring nodes is the node whose neighboring nodes are the first node in the common set. 1 node For the first The node and the first in the co-set The weight of the edge formed by the nodes. For the first The node and the first in the co-set The weights of the edges formed by the nodes. In this embodiment, when calculating the score of each node becoming the center node, both node centrality and specificity are considered, ensuring that users within the same class are close while users between classes are far apart, resulting in high accuracy. Furthermore, the calculation of node centrality also considers the distance between users and the amount of tasks generated by each user. With a limited number of drones, priority is given to users who generate a large amount of tasks, thus improving the subsequent data calculation workload of the entire system.

[0063] Step 3.4: Select the node with the highest centrality from all nodes. Centrality, and the former The nodes corresponding to the centrality are taken as the central nodes, and then we obtain One central node.

[0064] Step 4: Determine the central node associated with each non-central node, and... Each node is divided into Each module contains a central node and non-central nodes associated with that central node, utilizing... Module calculation The location of each drone is determined and updated iteratively until the coordinates of each drone remain stable. The final stable coordinates of each drone are then used as the final deployment location of the drone.

[0065] Step 4.1: Determine the central node associated with each non-central node, including:

[0066] For each non-central node, the central node with the highest Jaccard similarity to the current non-central node is selected as the associated central node, and then the scale-free distance-weighted network is further processed. The nodes were divided into Categories (i.e.) (each module)

[0067] Step 4.2, using Module calculation The location of each drone is recorded and updated iteratively (using Weiszfeld iteration) until the coordinates of each drone stabilize. The final stable coordinates of each drone are then used as the final deployment location of the drones, including:

[0068] Each module and Each drone corresponds to a specific drone;

[0069] Step 4.2.1: For each module, calculate the average coordinates of all nodes contained in each module to obtain the initial position of each UAV. ),at this time ;

[0070] Step 4.2.2: Update and iterate the positions of each UAV, and the update formula is as follows:

[0071] (1);

[0072] Formula (1) is equivalent to:

[0073] ;

[0074] ;

[0075] in,

[0076] ;

[0077] in, For the first The next update iteration uses the current drone's x-coordinate or y-coordinate (where... For the first The next update iteration will be based on the current drone's x-coordinate. For the first The next update iteration uses the current drone's ordinate. For the first In the next update iteration, the module corresponding to the current drone is the [number]th [item]. The x-coordinate or y-coordinate of each node (where For the first In the next update iteration, the module corresponding to the current drone is the [number]th [item]. The x-coordinate of each node, For the first In the next update iteration, the module corresponding to the current drone is the [number]th [item]. (the y-coordinate of each node) This represents the total number of nodes in the module corresponding to the current drone. For the first In the next update iteration, the module corresponding to the current drone is the [number]th [item]. The distance between each node and the current drone. For the first The next update iteration includes all nodes in the module corresponding to the current drone and the... The reciprocal of the sum of the distances of the current drone during the next update iteration;

[0078] Step 4.2.3: Repeat step 4.2.2 until the Euclidean distance between the positions of each UAV in two consecutive update iterations is less than the set threshold. If the update iteration stops, the location of each drone corresponding to the last update iteration will be taken as the final deployment location of the drone.

[0079] Figure 3 This diagram illustrates how to use users in a scenario to locate and deploy drones, where green dots represent users and red dots represent drones.

[0080] This method is compared with two existing methods (such as clustering algorithms) in the present technology. Figure 4 The K-means algorithm and the frog-leapfrog algorithm were used to compare the percentage of users covered by drones. Figure 4 It can be seen that this method covers a higher percentage of users with drones, thus making the drone positioning and deployment more accurate.

[0081] Compared to traditional heuristic algorithms, such as the Frog Leaping Algorithm (SFLA), this method is not limited by the initial node selection, has less dependence on parameters, and can maintain convergence stability even with a large number of nodes. Compared to clustering algorithms, such as K-Means, the centrality and specificity calculations in this method can be reused; the corresponding central nodes can be added or removed based on the increase or decrease of nodes. If the drone coverage area changes, the steps before pruning can be reused, and pruning can be performed according to the new coverage area, making it flexible and adaptable to changing environments. For example, if the drone service level distance changes due to environmental reasons, the network can be re-established based on the distance between users. When the number of drones increases or decreases, this method can quickly adjust to the new number of drones using the already determined scores, demonstrating strong adaptability.

[0082] This scale-free attribute-based task scheduling-oriented UAV positioning and deployment method constructs a scale-free distance-weighted network, determines the central node of the scale-free distance-weighted network and the non-central nodes associated with each central node, uses each module to determine the position of each UAV, and uses Weiszfeld to iteratively update, thereby achieving the final positioning and deployment of each UAV.

[0083] This method considers both node centrality and specificity when calculating the score for each node to become a central node, ensuring that users within the same class are close while users between classes are far apart, resulting in high accuracy. Furthermore, the calculation of node centrality takes into account both the distance between users and the amount of tasks generated by each user. When the number of drones is limited, priority is given to users who generate a large amount of tasks, thereby improving the subsequent data calculation workload of the entire system.

[0084] It should be understood that, although Figure 1 The steps in the flowchart are shown sequentially as indicated by the arrows, but these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified in this document, there is no strict order in which these steps are executed, and they can be performed in other orders. Figure 1 At least some of the steps in the process may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily executed at the same time, but may be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but may be executed in turn or alternately with other steps or at least some of the sub-steps or stages of other steps.

[0085] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0086] The embodiments described above are merely illustrative of several implementations of the present invention, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of the invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these modifications and improvements all fall within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the appended claims.

Claims

1. A method for UAV positioning and deployment based on scale-free attribute-oriented task scheduling, characterized in that: The scale-free attribute-based task scheduling-oriented UAV positioning and deployment method is applied in scenarios involving multiple UAVs and multiple users, wherein: The scene contains A drone and Individual users, and drones provide edge computing services to users; Will A scale-free distance-weighted network is constructed for each user, with each user as a node in the scale-free distance-weighted network. The edges connecting any two nodes are pruned, and the remaining edges form the edges of the scale-free distance-weighted network. The weight of each edge is then calculated. Determine the scale-free distance-weighted network One central node; Determine the central node associated with each non-central node, and then... Each node is divided into Each module contains a central node and non-central nodes associated with that central node, utilizing... Module calculation The location of each drone is determined and updated iteratively until the coordinates of each drone remain stable. The final stable coordinates of each drone are then used as the final deployment location of the drone. Among them, the determination of the scale-free distance-weighted network The central nodes include: Normalize the amount of tasks generated by each node; The centrality of each node is obtained by weighting and summing the degree of each node with the normalized task amount of the corresponding node. The degree of each node is the sum of the edge weights formed by all other nodes in the scale-free distance weighted network and the current node. Calculate the score by which each node becomes the central node based on the specific characteristics of each node; Select the node with the highest centrality among all nodes. Centrality, and the former The nodes corresponding to the centrality are taken as the central nodes, and then we obtain One central node.

2. The UAV positioning and deployment method based on scale-free attribute-oriented task scheduling as described in claim 1, characterized in that: Within the same time slot, each user's location is fixed; between different time slots, each user moves. .

3. The UAV positioning and deployment method based on scale-free attribute-oriented task scheduling as described in claim 1, characterized in that: If the Euclidean distance between the two nodes corresponding to the connecting edge is greater than a preset value, the current connecting edge will be pruned.

4. The UAV positioning and deployment method based on scale-free attribute-oriented task scheduling as described in claim 1, characterized in that: The formula for calculating the edge weights of a scale-free distance-weighted network is as follows: ； in, For the first The node and the first The weight of the edge formed by the nodes. Let be the set of all nodes in a scale-free distance-weighted network. For the first The node and the first Euclidean distance between nodes The maximum distance at which drones can provide edge computing services to users.

5. The UAV positioning and deployment method based on scale-free attribute-oriented task scheduling as described in claim 1, characterized in that: The formula for calculating the score of each node as the center node is as follows: ； in, ； in, For the first The score of each node becoming the central node For the first Node specificity For the first The node and the first Jaccard similarity of nodes, For the first Centrality of each node The set consisting of the centrality of all nodes. This indicates taking the maximum value. This indicates taking the minimum value.

6. The UAV positioning and deployment method based on scale-free attribute-oriented task scheduling as described in claim 1, characterized in that: For each non-central node, select the central node with the highest similarity to the current non-central node Jaccard as the central node associated with the current non-central node.

7. The UAV positioning and deployment method based on scale-free attribute-oriented task scheduling as described in claim 1, characterized in that: Each module and Each drone corresponds to a specific drone; For each module, the coordinates of all nodes contained in each module are averaged to obtain the initial position of each UAV; The positions of each drone are updated iteratively until the Euclidean distance between the positions of two consecutive updates is less than a set threshold. If the update iteration stops, the position of each drone corresponding to the last update iteration will be taken as the final deployment location of the drone, and the formula for updating the position of each drone is: ； in, For the first The next update iteration changes the current drone's x-coordinate or y-coordinate. For the first In the next update iteration, the module corresponding to the current drone is the [number]th [item]. The x-coordinate or y-coordinate of each node, This represents the total number of nodes in the module corresponding to the current drone. For the first In the next update iteration, the module corresponding to the current drone is the [number]th [item]. The distance between each node and the current drone. For the first The next update iteration includes all nodes in the module corresponding to the current drone and the... The reciprocal of the sum of the distances of the current drone during the next update iteration.

Images