Distributed robust topology control method and system for cluster unmanned aerial vehicles

Abstract

The invention provides a distributed robust topology control method and system for cluster unmanned aerial vehicles, and the method comprises the steps: carrying out the construction of a midpoint coverage circle with a node as a center for each node of a cluster unmanned aerial vehicle network, and judging potential key nodes; acquiring local topology through k times of neighbor information exchange for potential key nodes, generating a simplified block tree model based on depth-first search, traversing connected components by taking the simplified block tree model as a root node to form block point sets, and judging that the root node is a key node when the number of the block point sets to which the root node belongs is greater than 1; and constructing an enhanced edge candidate set which is arranged according to a cost ascending order by the determined key nodes, selecting the edge with the minimum cost to communicate different blocks, sending a moving instruction to controlled nodes in k hops to establish a communication link, and finally enabling the network topology to meet vertex 2-communication. According to the method, the key nodes in the network are detected, and the key nodes are eliminated in a node moving mode, so that the stable and reliable unmanned aerial vehicle cluster network topology is constructed at low cost.

Description

Technical Field

[0001] The present invention belongs to the field of unmanned aerial vehicle (UAV) network communication technology, and in particular relates to a distributed robust topology control method and system for clustered UAVs. Background Art

[0002] At present, drone swarm technology has become an important development trend in the drone field. That is, the use of multiple drones carrying various equipment to form a network can greatly improve the capabilities of payload, perception, and information processing, thereby achieving large-scale, high-efficiency, and multi-task operation goals.

[0003] For drone swarms, a stable and reliable network topology is crucial, as it determines the overall communication quality of the network and the collaborative capabilities of the drones. To build the required topology, existing research has proposed a variety of topology control methods, using a variety of techniques such as adding relay nodes, adjusting transmission power, modifying node positions, and joint optimization. Furthermore, given that drones need to act as relay nodes in the network, each one impacts the connectivity of the entire network. If a drone malfunctions or is interfered with, it could disrupt the communication paths between other drones. Therefore, some research has also proposed detection methods for key nodes in the network that may affect the topology.

[0004] However, these methods still have some shortcomings when applied to drone cluster networks: (1) Most topology control algorithms are centralized, and their execution depends on global topology information. This feature limits their application in large-scale drone swarm networks. Some distributed algorithms have difficulty in ensuring the control effect of the global network topology.

[0005] (2) Most existing topology control algorithms are designed for two-dimensional networks. The three-dimensional flight characteristics of UAVs are not fully considered.

[0006] (3) Existing research has not yet addressed the balance between algorithm performance and overhead. Summary of the Invention

[0007] The purpose of the present invention is to provide a distributed robust topology control method and system for swarm drones to solve the above problems.

[0008] To achieve the above object, the present invention adopts the following technical solutions: In a first aspect, the present invention provides a distributed robust topology control method for swarm drones, comprising: For each node in the swarm drone network, a midpoint coverage circle with the node as the center is constructed. If there is a non-unique node in the circular area, the node is judged as a non-key node, or if the node neighbor degree is 1, the node is judged as a non-key node. The remaining undetermined nodes are marked as potential key nodes. For potential key nodes, local topology is obtained through k-time neighbor information exchange, and a simplified block tree model is generated based on depth-first search. The root node is used as the root node to traverse the connected components to form a block point set. When the number of block point sets to which the root node belongs is greater than 1, it is determined to be a key node. The determined key nodes construct a set of reinforcement edge candidates sorted in ascending order of cost, select the minimum cost edge to connect different blocks, and send movement instructions to the controlled nodes within k hops to establish communication links, ultimately making the network topology meet vertex 2-connectivity.

[0009] Furthermore, for each node in the swarm drone network, a midpoint coverage circle with a radius of the node as the center is constructed. When there is a non-unique node in the circular area, the node is determined to be a non-key node, or the node neighbor degree is 1, and the remaining undetermined nodes are marked as potential key nodes, including: Each node executes a detection algorithm based on the midpoint covering circle, including: For the graph Any node n, n and its neighboring nodes form a graph A local subgraph of ; In the figure In the graph, if there is a pair of nodes i and j that are not adjacent nodes, they can only communicate through node n, and the shaded area A is the intersection area of the coverage of nodes i and j; if there is only node n in the shaded area A, then n is the If there are other nodes in the shaded area A, then n is not a key node of the graph. A key node, if node n is not a graph The key node in the graph is not n. The key nodes in .

[0010] A circular area is constructed with node n as the center and the average value of d as the radius, which is called the midpoint covering circle. d is the distance between m and n, and m is the midpoint of the line connecting nodes i and j. If node n is not the only node in the area, node n is determined to be a non-critical node; if node n is not the only node in the area, node n is determined to be a non-critical node. For nodes with a neighboring node degree of 1, they are not critical nodes.

[0011] Furthermore, is the radius of the midpoint covering circle, and we get

[0012] make

[0013] And suppose for The joint probability density function of

[0014]

[0015] Where r is the coverage radius of the nodes in the network; According to the position relationship between nodes

[0016] in ; When a is determined, that is, the position of node i is fixed

[0017]

[0018] in ; If the positions of nodes i and j are fixed, then determine The value of The values of follow a uniform distribution:

[0019] in ; When node j is located at the intersection of the coverage sections of nodes n and i, The value of reaches the minimum value, then , The minimum value of ; Substituting formulas (4), (5) and (6) into (3), we get

[0020]

[0021] The average value of d in three-dimensional space is obtained by formula (7): The approximate minimum value of constructs a node with node n as the center. is a spherical area with radius .

[0022] Furthermore, the method of acquiring local topology by exchanging k neighbor information for potential key nodes, generating a simplified block tree model based on depth-first search, taking itself as the root node to traverse connected components to form a block point set, and determining that a root node is a key node when the number of block point sets to which it belongs is greater than 1, includes: The key nodes are further identified by the detection algorithm based on k-hop local topology: If a node is not determined as a non-critical node by MCCDA, it will be further tested through KLTDA. In this stage, node v exchanges k Hello messages with its neighbors to obtain a k-hop subgraph; then, node v uses a block tree model to simplify the k-hop subgraph and detect whether it is a k-hop critical node.

[0023] Furthermore, the block tree model is further simplified into a block tree model of a single key node. , each connected component including the key node is equivalent to a block, and the node The star-structured block tree model; A simplified block tree model of the node local topology is generated using depth-first search; First, the node itself is used as the root node of the block tree. Starting from the root node, any unvisited neighbor node of the current node is always visited, and the passed vertices are marked as visited and recorded in a block point set; this process is continued until all the nodes of a branch have been visited, and then the process terminates; at this time, vertex backtracking is performed, backtracking to the root node to continue searching other branches; when all nodes have been visited, the search ends; after the block tree is constructed, each block point set represents a block to which this node belongs, and the nodes in the set represent the nodes in the block; when the number of block point sets of the root node is greater than 1, that is, it belongs to at least two different blocks, then the node is determined to be a key node.

[0024] Furthermore, the determined key nodes construct a set of enhanced edge candidates sorted in ascending order of cost, select the minimum cost edge to connect different blocks, and send movement instructions to the controlled nodes within k hops to establish communication links, ultimately making the network topology meet vertex 2-connectivity, including: When a graph satisfies 2-connectivity, the corresponding network topology is robust. The vertex 2-connectivity graph hardening problem is defined as follows: Given an initial simple connected graph , construct a complete graph using the vertices in the graph , edge set Represents the set of all edges that can be used to enhance the connectivity properties of the graph, called the enhanced edge set. Each edge in the network is assigned a unique cost or weight greater than zero. ; Use the node movement method to add the strengthening edge operation, and set the cost of the strengthening edge to be the Euclidean distance between the two vertices connected by the edge minus the maximum communication distance of the node, that is, , is the Euclidean distance between any two vertices; the optimization goal is to determine the candidate subset of strengthened edges , so that the graph with the reinforcement edge added For vertices to be 2-connected, the cost of all strengthening edges and is the smallest; Introducing node mobility constraints: Initially, the node and adjacent, the distance is If the node To another node Move, then when and The distance is the maximum communication distance When , get the node Towards Maximum distance moved ; From the cosine theorem, we get

[0025] The solution is

[0026] in express and The angle between the two vectors, and because ,

[0027] so

[0028] For other neighboring nodes of a node, the corresponding maximum moving distance can be obtained in the same way. In summary, the mobility constraint of a node is defined as: In either direction To ensure that all its neighboring nodes The maximum distance that can be moved is based on the premise of connectivity, as shown in the formula:

[0029] in express The maximum moving distance set calculated by formula (12) with each neighboring node.

[0030] Furthermore, when selecting a reinforcement edge, it is ensured that the cost of the reinforcement edge cannot be greater than the sum of the maximum movable distances of its two endpoints, that is, ; A reinforcement edge search algorithm based on a simplified block tree model is used to search for reinforcement edge solution sets. , the specific steps are as follows: (1) Initialization: Initialize the reinforcement edge set is an empty set; (2) Constructing the search space: Traverse all block graphs based on the simplified block tree model of key nodes, search for the shortest strengthening edge between any two different blocks, and the strengthening edge must satisfy the node mobility constraint, and then add it to the strengthening edge set And sort them in ascending order of weight value; (3) Greedy selection: Select the edge with the smallest weight from the set of strengthened edges, and use the union-check method to determine whether the two blocks to which the edge belongs are already connected; if not, add it to the solution set. ; If it is already connected, abandon the enhanced edge; finally remove the edge from the enhanced edge set; (4) Termination condition: When the solution set The number of edges in or enhanced edge set When it is an empty set, the algorithm terminates and returns the solution set .

[0031] In a second aspect, the present invention provides a distributed robust topology control system for swarm drones, comprising: The potential node judgment module is used to execute a midpoint coverage circle with a radius of 1 for each node in the swarm drone network. When there is a non-unique node in the circular area, it is judged as a non-key node, or the node neighbor degree is 1. The remaining undetermined nodes are marked as potential key nodes. The key node judgment module is used to obtain the local topology of potential key nodes through k-time neighbor information exchange, generate a simplified block tree model based on depth-first search, and use itself as the root node to traverse the connected components to form a block point set. When the number of block point sets to which the root node belongs is greater than 1, it is judged as a key node; The judgment output module is used to determine the key nodes obtained to construct a set of reinforcement edge candidates arranged in ascending order of cost, select the minimum cost edge to connect different blocks, and send movement instructions to the controlled nodes within k hops to establish communication links, ultimately making the network topology meet the vertex 2-connectivity.

[0032] In a third aspect, the present invention provides a computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the distributed robust topology control method for clustered drones when executing the computer program.

[0033] In a fourth aspect, the present invention provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, it implements the steps of the distributed robust topology control method for cluster drones.

[0034] Compared with the prior art, the present invention has the following technical effects: The topology control scheme proposed in this paper is based on k-hop local topology. Therefore, global topology information is not required, reducing the control overhead of information exchange. This is very beneficial for large-scale and dynamic drone swarm networks. Furthermore, the proposed MCCDA can quickly identify non-critical nodes using 1-hop neighbor information with very low error, further reducing control overhead. In contrast, centralized algorithms such as block-moving and compaction require global topology information and rely on global movement strategies. Therefore, the control overhead of the proposed scheme is lower than that of block-moving and compaction algorithms.

[0035] The success rate of building a robust topology using the present invention is similar to that of the block-moving algorithm and far superior to that of the compression algorithm. Furthermore, the present invention has an even greater success rate in large-scale networks. This is because the present invention and the block-moving algorithm share the same underlying principle: connecting the blocks to which key nodes belong (the differences are discussed below). However, the compression algorithm eliminates key nodes by simply moving all nodes to the center of the network, an operation that fails to ensure the robustness of the resulting topology.

[0036] Although the present invention and the block moving algorithm have similar success rates, the block moving algorithm has a much larger moving overhead. The reason is that the block moving algorithm eliminates all key nodes in the network by moving all nodes within a block from a global perspective. In contrast, the present invention only moves a pair of nodes from a local perspective, and models the problem as a mathematical problem (minimum spanning tree (MST)), and provides a good solution, which greatly reduces the moving overhead. The compression algorithm needs to move all nodes in the network, resulting in huge moving overhead. Therefore, the present invention has better moving overhead control capabilities and is of great significance for energy-constrained drone networks. Under fixed density, the success rate of all algorithms decreases as the number of nodes increases. However, the present invention maintains good performance in large-scale networks, and its advantages over block-moving and compression algorithms become even more pronounced. This is because block-moving and compression algorithms are based on global topology. In large-scale networks, they consume more control and movement overhead. Furthermore, as network size increases, it becomes increasingly difficult to eliminate critical nodes by moving all nodes within a block. In contrast, increasing network size (at the same density) has less impact on local topology. Because the present invention is implemented based on local topology, it still performs well. However, in large-scale networks, the probability of no feasible solution to the minimum spanning tree problem increases, and the success rate may decrease, but it still significantly outperforms block-moving and compression algorithms. Therefore, the present invention has greater scalability, making it more practical in large-scale network applications. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 Neighbor node location relationship.

[0038] Figure 2 Diagram of a simplified block tree model of a graph.

[0039] Figure 3 Flowchart of the present invention.

[0040] Figure 4 Comparison chart of test cases of the present invention.

[0041] Figure 5 A comparison chart of the success rates of the present invention and the comparative topology control algorithms.

[0042] Figure 6 Comparison of the total moving distances of the present invention and the comparative topology control algorithm.

[0043] Figure 7 Comparison diagram of node movement ratios of the present invention and the comparative topology control algorithm. DETAILED DESCRIPTION

[0044] The present invention will be further described below with reference to the accompanying drawings: In embodiment 1, the present invention provides a distributed robust topology control method for swarm drones, comprising: For each node in the swarm drone network, a midpoint coverage circle with the node as the center is constructed. If there is a non-unique node in the circular area, the node is judged as a non-key node, or if the node neighbor degree is 1, the node is judged as a non-key node. The remaining undetermined nodes are marked as potential key nodes. For potential key nodes, local topology is obtained through k-time neighbor information exchange, and a simplified block tree model is generated based on depth-first search. The root node is used as the root node to traverse the connected components to form a block point set. When the number of block point sets to which the root node belongs is greater than 1, it is determined to be a key node. The determined key nodes construct a set of reinforcement edge candidates sorted in ascending order of cost, select the minimum cost edge to connect different blocks, and send movement instructions to the controlled nodes within k hops to establish communication links, ultimately making the network topology meet vertex 2-connectivity.

[0045] This method detects key nodes in the network and eliminates them by moving nodes, thereby building a stable and reliable drone cluster network topology at low cost.

[0046] In Example 2, the present invention provides a distributed robust topology control method for swarm drones, specifically comprising: S1: Use adjacent geometric relationships to quickly screen potential key nodes; S2: Actively detect potential key nodes based on k-hop local topology information; S3: Key nodes use the enhanced edge search algorithm to find the optimal movement plan and build a robust network topology by controlling node movement.

[0047] Specifically, in step S1, each node executes the Midpoint Coverage Circle Based Detection Algorithm (MCCDA) to quickly identify whether it is a non-critical node using its one-hop neighbor information. If it is determined to be a non-critical node, no further action is required. Otherwise, the node proceeds to the next step.

[0048] For the graph Any node n, n and its neighboring nodes form a graph A local subgraph of In Figure In , if there is a pair of nodes i and j that are not adjacent nodes, they can only communicate through node n, such as Figure 1 As shown, the shaded area A is the intersection area of the coverage areas of nodes i and j.

[0049] according to Figure 1 It can be concluded that if there is only node n in the shaded area A, then n is the graph If there are other nodes in the shaded area A, then n is not a key node of the graph. Assume that m is the midpoint of the line connecting nodes i and j, then m is located in the shaded area A. If a region can be constructed so that the graph The midpoints of any pair of non-adjacent nodes in the graph are located in the constructed region with high probability. If node n is not the only node in the region, then node n is not a node in the graph. If node n is not a key node of the graph The key node in the graph is not n. The key nodes in .

[0050] Therefore, a circular area is constructed with node n as the center and the average value of d (the distance between m and n) as the radius, which is called the Midpoint Coverage Circle (MCC). If node n is not the only node in the area, node n is determined to be a non-critical node.

[0051] set up is the radius of the midpoint covering circle, according to Figure 1 , we can get

[0052] make

[0053] And suppose for The joint probability density function of

[0054]

[0055] Where r is the coverage radius of the nodes in the network.

[0056] Because node i can be located anywhere within the coverage of node n, according to the positional relationship between nodes, we can get

[0057] in .

[0058] When a is determined, that is, the position of node i is fixed, node j can only be located in area B, such as Figure 1 shown. Figure 1 for Figure 1 From this we can infer that

[0059]

[0060] in .

[0061] If the positions of nodes i and j are fixed, we can determine The value of The value of follows a uniform distribution, such as Figure 1 The shaded area is shown. Therefore,

[0062] in .

[0063] Obviously, in Figure 1 In , when node j is located at the intersection of the coverage sections of node n and node i, The value of reaches the minimum value, then .therefore, The minimum value of .

[0064] Substituting formulas (4), (5) and (6) into (3), we get

[0065]

[0066] The average value of d in three-dimensional space can be obtained by formula (7): The approximate minimum value of , from which we can construct a node with n as the center, A spherical region with a radius of . If node n is not the only node in this region, then node n is not considered a key node. Furthermore, a node with a neighboring node degree of 1 cannot be a key node in the network, so nodes with a neighboring node degree of 1 can also be considered non-key nodes.

[0067] Specifically, in step S2, for undetermined nodes, it is necessary to obtain k-hop local topology information and further identify whether they are key nodes through a k-hop local topology based detection algorithm (KLTDA).

[0068] If a node is not identified as a non-critical node by MCCDA, it is further tested using KLTDA. In this phase, node v exchanges k Hello messages with its neighbors to obtain a k-hop subgraph. Node v then uses a block tree model to simplify the k-hop subgraph and detect whether it is a k-hop critical node.

[0069] The present invention further simplifies the block tree model into a block tree model of a single key node, such as Figure 2 As shown. For the node , each connected component including the key node is equivalent to a block, so that the node The star-structured block tree model.

[0070] A node v is considered a key node in a graph G only if it belongs to at least two different blocks of G. In other words, whether a node in a network is a key node is determined by the number of blocks it belongs to. Therefore, after a node exchanges information with its neighbors to obtain a k-hop local topology, we use depth-first search (DFS) to generate a simplified block tree model of the node's local topology.

[0071] First, the node itself is used as the root node of the block tree. Starting from the root node, any unvisited neighbor node of the current node is always visited, and the vertices passed are marked as visited and recorded in a block point set. This process continues until all the nodes of a branch have been visited, and the process terminates. At this time, vertex backtracking is performed, backtracking to the root node to continue searching other branches. When all nodes have been visited, the search ends. After the block tree is constructed, each block point set represents a block to which this node belongs, and the nodes in the set represent the nodes in the block. When the number of block point sets of the root node is greater than 1, that is, it belongs to at least two different blocks, the node is determined to be a key node.

[0072] Specifically, in step S3, based on the key node detection algorithm, if any node identifies itself as a key node, it executes the enhanced edge search algorithm proposed in this invention to find the optimal drone movement plan. The node then sends control information to its corresponding neighbor nodes within the k-hop local topology, enabling them to establish communication links with the corresponding nodes through movement. This adds enhanced edges and eliminates key nodes, ultimately achieving a robust network topology.

[0073] When a graph satisfies 2-connectivity, the corresponding network topology is robust. The problem of strengthening a vertex 2-connectivity graph is to directly add "strengthening edges" between different vertices in a simple connected graph to eliminate key nodes and achieve vertex 2-connectivity. The optimization goal is to minimize the total cost of the set of strengthened edges, which is defined as follows: Given an initial simple connected graph , construct a complete graph using the vertices in the graph , edge set Represents the set of all edges that can be used to enhance the connectivity properties of the graph, called the enhanced edge set. Each edge in the network is assigned a unique cost or weight greater than zero. In view of the maneuverability of drones, the present invention chooses to use the node movement method to add the enhanced edge operation, so the cost of the enhanced edge is set to the Euclidean distance between the two vertices connected by the edge minus the maximum communication distance of the nodes, that is, , is the Euclidean distance between any two vertices. The optimization goal is to determine the candidate subset of the strengthened edge , so that the graph with the reinforcement edge added For vertices to be 2-connected, the cost of all strengthening edges and is the minimum.

[0074] In the algorithm, each key node needs to determine the available reinforcement edge set based on the local topology and neighbor node information. If we construct the enhanced edge set directly by constructing the complete graph , the data space occupied will be huge and the execution speed of the algorithm will be reduced. Therefore, it is necessary to strengthen the edge set Pruning is performed to narrow the search space of feasible solutions, reduce storage resource usage, and improve algorithm execution speed.

[0075] In order to avoid affecting the structure of the initial network topology and optimize the edge set during the node mobility control process, the node mobility constraint is introduced.

[0076] Initially, the node and adjacent, the distance is If the node To another node Move, then when and The distance is the maximum communication distance When , get the node Towards Maximum distance moved .

[0077] From the cosine theorem, we can get

[0078] The solution is

[0079] in express and The angle between the two vectors, and because ,

[0080] so

[0081] For other neighboring nodes of a node, the corresponding maximum moving distance can be obtained in the same way. In summary, the mobility constraint of a node is defined as: In either direction To ensure that all its neighboring nodes The maximum distance that can be moved is given by the formula (12).

[0082]

[0083] in express The maximum moving distance set calculated by formula (12) with each neighboring node.

[0084] Therefore, when selecting a reinforcement edge, it should be ensured that the cost of the reinforcement edge is not greater than the sum of the maximum movable distances of its two endpoints, that is, .

[0085] Furthermore, for any key node, it is necessary to strengthen the edge set Find n-1 reinforcement edges in the network so that n different blocks are connected. Therefore, the present invention proposes a simplified block tree based reinforcement edge search algorithm (SBT-RESA) to search for the reinforcement edge solution set. , the specific steps are as follows: (1) Initialization: Initialize the reinforcement edge set Is an empty set.

[0086] (2) Constructing the search space: Traverse all block graphs based on the simplified block tree model of key nodes, search for the shortest strengthening edge between any two different blocks, and the strengthening edge must satisfy the node mobility constraint, and then add it to the strengthening edge set And sort them in ascending order of weight value.

[0087] (3) Greedy selection: Select the edge with the smallest weight from the set of strengthened edges, and use the union-check method to determine whether the two blocks to which the edge belongs are already connected. If not, add it to the solution set. If it is already connected, abandon the enhanced edge. Finally, remove the edge from the enhanced edge set.

[0088] (4) Termination condition: When the solution set The number of edges in or enhanced edge set When it is an empty set, the algorithm terminates and returns the solution set .

[0089] Finally, the proposed topology control scheme adds strengthening edges to the network topology through node movement, achieving a robust network topology. After the strengthening edge search algorithm determines the set of strengthening edges to be added, key nodes use this set to send control information to the nodes that need to be moved within their k-hop local topology, thereby establishing corresponding communication links between the nodes.

[0090] In another embodiment of the present invention, a distributed robust topology control system for swarm drones is provided, which can be used to implement the above-mentioned distributed robust topology control method for swarm drones. Specifically, the system includes: The potential node judgment module is used to execute a midpoint coverage circle with a radius of 1 for each node in the swarm drone network. When there is a non-unique node in the circular area, it is judged as a non-key node, or the node neighbor degree is 1. The remaining undetermined nodes are marked as potential key nodes. The key node judgment module is used to obtain the local topology of potential key nodes through k-time neighbor information exchange, generate a simplified block tree model based on depth-first search, and use itself as the root node to traverse the connected components to form a block point set. When the number of block point sets to which the root node belongs is greater than 1, it is judged as a key node; The judgment output module is used to determine the key nodes obtained to construct a set of reinforcement edge candidates arranged in ascending order of cost, select the minimum cost edge to connect different blocks, and send movement instructions to the controlled nodes within k hops to establish communication links, ultimately making the network topology meet the vertex 2-connectivity.

[0091] The module division in the embodiments of the present invention is illustrative and represents only one logical functional division. In actual implementation, other division methods may be used. Furthermore, the functional modules in various embodiments of the present invention may be integrated into a single processor, exist physically as separate modules, or two or more modules may be integrated into a single module. The integrated modules may be implemented in either hardware or software functional modules.

[0092] In another embodiment of the present invention, a computer device is provided, which includes a processor and a memory, wherein the memory is used to store a computer program, the computer program includes program instructions, and the processor is used to execute the program instructions stored in the computer storage medium. The processor may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. It is the computing core and control core of the terminal, which is suitable for implementing one or more instructions, specifically suitable for loading and executing one or more instructions in the computer storage medium to implement the corresponding method flow or corresponding function; the processor described in the embodiment of the present invention can be used for the operation of a distributed robust topology control method for clustered drones.

[0093] In another embodiment of the present invention, a storage medium is provided, specifically a computer-readable storage medium (Memory). The computer-readable storage medium is a memory device in a computer device, used to store programs and data. It is understood that the computer-readable storage medium herein may include both built-in storage media in the computer device and, of course, extended storage media supported by the computer device. The computer-readable storage medium provides storage space that stores the terminal's operating system. Furthermore, the storage space also stores one or more instructions suitable for being loaded and executed by a processor. These instructions may be one or more computer programs (including program code). It should be noted that the computer-readable storage medium herein may be high-speed RAM memory or non-volatile memory, such as at least one disk drive. The processor may load and execute the one or more instructions stored in the computer-readable storage medium to implement the corresponding steps of the distributed robust topology control method for swarm drones described in the above-mentioned embodiment.

[0094] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Thus, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0095] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowcharts and / or block diagrams, as well as combinations of processes and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowcharts and / or block diagrams. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0096] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0097] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0098] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in the field should understand that the specific implementation methods of the present invention can still be modified or replaced by equivalents. Any modification or equivalent replacement that does not depart from the spirit and scope of the present invention should be covered by the scope of protection of the claims of the present invention.

Claims

1. A distributed robust topology control method for swarm drones, characterized by: include: For each node in the swarm drone network, a midpoint coverage circle with the node as the center is constructed. If there is a non-unique node in the circular area, the node is judged as a non-key node, or if the node neighbor degree is 1, the node is judged as a non-key node. The remaining undetermined nodes are marked as potential key nodes. For potential key nodes, local topology is obtained through k-time neighbor information exchange, and a simplified block tree model is generated based on depth-first search. The root node is used as the root node to traverse the connected components to form a block point set. When the number of block point sets to which the root node belongs is greater than 1, it is determined to be a key node. The determined key nodes construct a set of reinforcement edge candidates sorted in ascending order of cost, select the minimum cost edge to connect different blocks, and send movement instructions to the controlled nodes within k hops to establish communication links, ultimately making the network topology meet vertex 2-connectivity.

2. A distributed robust topology control method for swarm drones according to claim 1, characterized in that: For each node in the swarm drone network, a midpoint coverage circle with a radius of the node as the center is constructed. When there is a non-unique node in the circular area, the node is determined to be a non-key node, or the node neighbor degree is 1. The remaining undetermined nodes are marked as potential key nodes, including: Each node executes a detection algorithm based on the midpoint covering circle, including: For the graph Any node n, n and its neighboring nodes form a graph A local subgraph of ; In the figure In the graph, if there is a pair of nodes i and j that are not adjacent nodes, they can only communicate through node n, and the shaded area A is the intersection area of the coverage of nodes i and j; if there is only node n in the shaded area A, then n is the If there are other nodes in the shaded area A, then n is not a key node of the graph. A key node, if node n is not a graph The key node in the graph is not n. The key nodes in A circular area is constructed with node n as the center and the average value of d as the radius, which is called the midpoint covering circle. d is the distance between m and n, and m is the midpoint of the line connecting nodes i and j. If node n is not the only node in the area, node n is determined to be a non-critical node; if node n is not the only node in the area, node n is determined to be a non-critical node. For nodes with a neighboring node degree of 1, they are not critical nodes.

3. A distributed robust topology control method for swarm drones according to claim 2, characterized in that: set up is the radius of the midpoint covering circle, and we get make And suppose for The joint probability density function of Where r is the coverage radius of the nodes in the network; According to the position relationship between nodes in ; When a is determined, that is, the position of node i is fixed in ; If the positions of nodes i and j are fixed, then determine The value of The values of follow a uniform distribution: in ; When node j is located at the intersection of the coverage sections of nodes n and i, The value of reaches the minimum value, then , The minimum value of ; Substituting formulas (4), (5) and (6) into (3), we get The average value of d in three-dimensional space is obtained by formula (7): The approximate minimum value of constructs a node with node n as the center. is a spherical area with radius .

4. The distributed robust topology control method for swarm drones according to claim 1, characterized in that: The method acquires the local topology of a potential key node by exchanging neighbor information k times, generates a simplified block tree model based on depth-first search, takes itself as the root node to traverse the connected components to form a block point set, and determines that a root node is a key node when the number of block point sets to which the root node belongs is greater than 1, including: The key nodes are further identified by the detection algorithm based on k-hop local topology: If a node is not determined as a non-critical node by MCCDA, it will be further tested through KLTDA. In this stage, node v exchanges k Hello messages with its neighbors to obtain a k-hop subgraph; then, node v uses a block tree model to simplify the k-hop subgraph and detect whether it is a k-hop critical node.

5. The distributed robust topology control method for swarm drones according to claim 4 is characterized in that: The block tree model is further simplified into a block tree model of a single key node. , each connected component including the key node is equivalent to a block, and the node The star-structured block tree model; A simplified block tree model of the node local topology is generated using depth-first search; First, take the node itself as the root node of the block tree. Starting from the root node, always visit any unvisited neighbor node of the current node, mark the passed vertices as visited, and record them in a block point set; This process continues until all nodes in a branch have been visited, and then the process terminates. At this point, backtrack to the root node and continue searching for other branches. When all nodes have been visited, the search ends. After the block tree is constructed, each block point set represents a block to which the node belongs, and the nodes in the set represent the nodes in the block. When the number of block point sets of the root node is greater than 1, that is, it belongs to at least two different blocks, the node is determined to be a key node.

6. The distributed robust topology control method for swarm drones according to claim 1, characterized in that: The key nodes obtained by the determination are used to construct a set of candidate enhanced edges sorted in ascending order of cost, and the minimum cost edge is selected to connect different blocks. Move instructions are sent to the controlled nodes within k hops to establish communication links, and finally the network topology satisfies vertex 2-connectivity, including: When the graph satisfies 2-connectivity, the corresponding network topology is robust. The vertex 2-connectivity graph enhancement problem is defined as follows: Given an initial simple connected graph , construct a complete graph using the vertices in the graph , edge set Represents the set of all edges that can be used to enhance the connectivity properties of the graph, called the enhanced edge set. Each edge in the network is assigned a unique cost or weight greater than zero. ; Use the node movement method to add the strengthening edge operation, and set the cost of the strengthening edge to be the Euclidean distance between the two vertices connected by the edge minus the maximum communication distance of the node, that is, , is the Euclidean distance between any two vertices; the optimization goal is to determine the candidate subset of strengthened edges , so that the graph with the reinforcement edge added For vertices to be 2-connected, the cost of all strengthening edges and is the smallest; Introducing node mobility constraints: Initially, the node and adjacent, the distance is If the node To another node Move, then when and The distance is the maximum communication distance When , get the node Towards Maximum distance moved ; From the cosine theorem, we get The solution is in express and The angle between the two vectors, and because , so For other neighboring nodes of a node, the corresponding maximum moving distance can be obtained in the same way. In summary, the mobility constraint of a node is defined as: In either direction To ensure that all its neighboring nodes The maximum distance that can be moved is based on the premise of connectivity, as shown in the formula: in express The maximum moving distance set calculated by formula (12) with each neighboring node.

7. The distributed robust topology control method for swarm drones according to claim 6, characterized in that: When selecting a reinforcement edge, ensure that the cost of the reinforcement edge cannot be greater than the sum of the maximum movable distances of its two endpoints, that is, ; A reinforcement edge search algorithm based on a simplified block tree model is used to search for reinforcement edge solution sets. , the specific steps are as follows: (1) Initialization: Initialize the reinforcement edge set is an empty set; (2) Constructing the search space: Traverse all block graphs based on the simplified block tree model of key nodes, search for the shortest strengthening edge between any two different blocks, and the strengthening edge must satisfy the node mobility constraint, and then add it to the strengthening edge set And sort them in ascending order of weight value; (3) Greedy selection: Select the edge with the smallest weight from the set of strengthened edges, and use the union-check method to determine whether the two blocks to which the edge belongs are already connected; if not, add it to the solution set. ; If it is already connected, abandon the enhanced edge; finally remove the edge from the enhanced edge set; (4) Termination condition: When the solution set The number of edges in or enhanced edge set When it is an empty set, the algorithm terminates and returns the solution set .

8. A distributed robust topology control system for swarm drones, characterized by: include: The potential node judgment module is used to execute a midpoint coverage circle with a radius of 1 for each node in the swarm drone network. When there is a non-unique node in the circular area, it is judged as a non-key node, or the node neighbor degree is 1. The remaining undetermined nodes are marked as potential key nodes. The key node judgment module is used to obtain the local topology of potential key nodes through k-time neighbor information exchange, generate a simplified block tree model based on depth-first search, and use itself as the root node to traverse the connected components to form a block point set. When the number of block point sets to which the root node belongs is greater than 1, it is judged as a key node; The judgment output module is used to determine the key nodes obtained to construct a set of reinforcement edge candidates arranged in ascending order of cost, select the minimum cost edge to connect different blocks, and send movement instructions to the controlled nodes within k hops to establish communication links, ultimately making the network topology meet the vertex 2-connectivity.

9. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the computer program, the steps of a distributed robust topology control method for cluster drones as described in any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of a distributed robust topology control method for clustered drones as described in any one of claims 1 to 7 are implemented.

Images