An adaptive dynamic topology invulnerability optimization method of a wireless self-organizing network

Abstract

The application discloses a kind of self-adapting dynamic topology anti-destroying optimization method of wireless self-organizing network, comprising the following steps: node detection and evaluation: the network is detected to each node, and the size of each node is evaluated by using split coefficient index to network anti-destroying influence, then based on the split coefficient index of node, sequentially preferred elimination node in network;Redundant relay node selection: if the algebraic connectivity of network topology is maximum after a certain communication relay node is deleted, the communication relay node is selected for topology optimization;Optimized deployment position determination: determine the target position of the communication relay node optimization deployment;Relay node moves and local adjustment: the communication relay node is moved and deployed to target position, and local adjustment is carried out according to the actual connectivity obtained in target position to effectively connect each connected branch.The application can enhance the connectivity between network nodes, thereby improving network anti-destroying.

Description

Technical Field

[0001] This invention relates to the field of wireless ad hoc network communication technology, and in particular to an adaptive dynamic topology resilience optimization method for wireless ad hoc networks. Background Technology

[0002] Wireless ad hoc networks (WANs) are rapidly deployable and flexibly assembled networks. Their network topology may constantly change due to node movement, communication environment, and other factors, and the communication relationships between nodes exhibit a multi-hop mesh connection. As nodes move within the network, the topology continuously changes due to altered neighbor relationships. Maintaining the connectivity and integrity of the entire network topology during these dynamic changes is a prerequisite for ensuring effective transmission of application information and a fundamental requirement for WAN applications.

[0003] To ensure network connectivity and integrity, network resilience is a key consideration in typical network design and planning. In wireless ad hoc networks, resilience refers to the network's ability to maintain communication availability even when nodes fail or are damaged. If, when some nodes fail or are damaged, one or more communication paths still exist between the remaining nodes, then the network possesses a certain degree of resilience. In real-world network applications, especially for military purposes, the failure of core critical nodes can lead to network fragmentation, compromising network connectivity and integrity.

[0004] In a multi-node, multi-hop relay wireless ad hoc network, the network's topology directly determines its resilience. This topology can be represented by a graph in graph theory, where connectivity is a crucial indicator of resilience. Therefore, a fundamental approach to improving network resilience is to increase the number of links between nodes, enriching the topological connections and preventing situations where communication between nodes is limited to a single, necessary path. Existing methods primarily focus on optimizing node transmit power to improve topology and resilience, but they fail to effectively address the connectivity issues that arise when links are interrupted due to node failure or neighboring nodes exceeding their maximum communication distance during node movement. Summary of the Invention

[0005] To address the issue of impaired network connectivity caused by node failures or link interruptions, this invention proposes an adaptive dynamic topology resilience optimization method for wireless ad hoc networks. By identifying weak points in the network topology and targeting key nodes affecting network connectivity, the method relocates and adjusts communication relay nodes in the network, thereby optimizing and changing the network topology, enhancing connectivity between network nodes, and thus improving network resilience.

[0006] The technical solution adopted in this invention is as follows:

[0007] An adaptive dynamic topology resilience optimization method for wireless ad hoc networks includes the following steps:

[0008] S1. Joint Detection and Evaluation: Joints in the network are detected, and the impact of each joint on the network's resilience is calculated and evaluated using the split coefficient index. Then, based on the split coefficient index of the joints, joints in the network are eliminated in order of preference.

[0009] S2. Redundant relay node selection: If the algebraic connectivity of the network topology is maximized after a certain communication relay node is deleted, then the communication relay node has the highest redundancy, and the communication relay node is selected for topology optimization.

[0010] S3. Determining the optimal deployment location: Determine the target location for the optimized deployment of the communication relay node, so that the moved communication relay node can effectively connect the connecting branches connected to the target node;

[0011] S4. Relay Node Movement and Local Adjustment: The communication relay node is moved and deployed to the target location, and local adjustments are made based on the actual connectivity obtained at the target location to effectively connect various connectivity branches, thereby eliminating key points in the network that affect resilience.

[0012] Furthermore, the network topology obtained by monitoring the communication relay nodes is represented by an undirected graph G = (V, E), where V is the set of nodes in the network and E is the set of connections between nodes in the network. If, for a node v ∈ V, after deleting node v and all edges associated with node v from the undirected graph G, the undirected graph G splits into two or more unconnected subgraphs, then node v is the key point of the undirected graph G.

[0013] Furthermore, in step S1, if N is found to exist in the network through keypoint detection... a If there are 1 key point, then the node set V is used. a Perform representation and |V a |=N a If, after removing key points, the undirected graph G is split into l connected components, the k-th connected component is represented as G. kThe number of nodes in each connected component are N1, N2, ..., N. l The formula for calculating the splitting coefficient index e is as follows:

[0014]

[0015] In the above formula, N is the number of ordinary nodes in the network, and N(N-1) represents the number of source-destination node communication pairs that the undirected graph G can support. k (N k -1) represents the connected component G. k The source-destination node communication pairs that can be supported, the larger the split coefficient index e, the greater the impact of the key point on the network's resilience.

[0016] Furthermore, for each set V... a For each keypoint, a splitting coefficient index e is calculated, and the keypoints are sorted according to the magnitude of the splitting coefficient index e. Subsequently, the keypoints will be eliminated in descending order of the splitting coefficient index e.

[0017] Furthermore, step S2, namely the selection of redundant relay nodes, includes the following sub-steps:

[0018] Step S201. Initialize the set

[0019] Step S202. Target each communication relay node Rv in the network one by one. i Delete node Rv i Next, check if any new keypoints have been generated in the network, where i = 1, 2, 3, ..., M; if no new keypoints have been generated, then node Rv is removed. i Add to set V c In the set V c It includes all redundant and movable communication relay nodes;

[0020] Step S203. Select set V c The communication relay node Rv with the least impact on network topology resilience k Priority scheduling is used to perform topology-optimized move deployments; for set V c Each node v in i ∈V c Delete node v i The network topology diagram after that is represented as G. i Calculate each graph G i The algebraic connectivity λ(G) i ), take the maximum value max{λ(G) i ), 1≤i≤|V c The node v corresponding to |} i As Rvk If multiple nodes have the same maximum value, then the node closest to the target location determined by the optimized deployment will be selected.

[0021] Furthermore, in step S203, the method for calculating algebraic connectivity includes:

[0022] If a network has n nodes, A(G) is the adjacency matrix representation of the network topology graph G:

[0023]

[0024] Among them, a ij =1 indicates node v i ,v j There is a link between them; otherwise, a ij =0;

[0025] Let D(G) denote the diagonal matrix formed by the node degrees of each node in graph G:

[0026]

[0027] Among them, node degree

[0028] Let L(G) denote the Laplacian matrix of graph G, and its calculation is expressed as:

[0029] L(G) = D(G) - A(G)

[0030] The eigenvalues ​​of the Laplace matrix L(G) are all non-negative real numbers, with the smallest eigenvalue being 0. Its n eigenvalues, after being sorted, are represented as 0 = λ₁ ≤ λ₂ ≤ ... ≤ λₙ. n The algebraic connectivity λ(G) of graph G is the second smallest eigenvalue of the Laplace matrix L(G), that is:

[0031] λ(G)=λ2

[0032] The larger the algebraic connectivity λ(G), the better the network connectivity and the better the network topology is.

[0033] Furthermore, if the key point v a The p connected components formed after deletion are WG1, WG2, ..., WG p Their node sets are represented as V1, V2, ..., V p Then step S3, which is the optimization of deployment location determination, includes the following sub-steps:

[0034] Step S301. Record the target location for optimized deployment of the communication relay node using R(x,y), and use D... mRecord the sum of the squared minimum distances to each connected component, and assign an initial value D. m =+∞;

[0035] Step S302. Sequentially from the connected component WG i Select one node from each of the following groups, and represent them as v1, v2, ..., v p Node v i The corresponding coordinate position is (x i ,y i ), where 1≤i≤p; compute node v i The geometric center position A(x) of the closed geometric shape formed d ,y d );

[0036] Step S303. Calculate the geometric center position A(x) d ,y d ) to each connected branch node v i The sum of the squares of the Euclidean distances is denoted by D:

[0037]

[0038] Step S304. Compare D and D m Calculate the size, take the minimum value, and update D. m Record, i.e., D m =min(D,D) m ); If there is D after the update m =D, then save D. m The target position corresponding to the value is the geometric center position A(x). d ,y d That is, R(x,y)=A(x) d ,y d );

[0039] Step S305. Repeat steps S302 to S304 until all connected components WG have been traversed. i The selection and combination of all nodes in the algorithm are calculated, and the final result R(x,y) is output, which is the target location for the optimized deployment of communication relay nodes.

[0040] Furthermore, in step S302, node v i The geometric center position A(x) of the closed geometric shape formed d ,y d )for:

[0041]

[0042] Furthermore, step S4, namely the relay node movement and local adjustment, includes the following sub-steps:

[0043] Step S401. Communication relay node Rv k Autonomously move to the target location;

[0044] Step S402. If it is not possible to achieve connection with all connected branches WG i If the nodes are connected, then the communication relay node Rv k To optimize the elimination of key object v a The location is locally moved and adjusted until it is connected to all connected branches WG. i All connections of nodes; in the minimum case, relay node Rv k Will move to the joint v a The location is the same as the joint point v. a Similarly, establish connections with all connected branches.

[0045] Furthermore, after completing the elimination of a key point, if there are still relevant nodes to be eliminated in the network, and there are also redundant and movable communication relay nodes, then steps S1 to S4 are executed iteratively to continuously optimize the network topology until all key points are eliminated.

[0046] The beneficial effects of this invention are as follows:

[0047] (1) This invention analyzes the network topology in real time, detects and evaluates each key point in the network, and adaptively selects the most suitable communication relay node to perform the mobile deployment operation to eliminate the key point. This expands the connectivity nodes and links of the key parts of the network, avoids the damage to the connectivity and integrity of the network caused by the failure of key nodes, and realizes the dynamic optimization and improvement of the network topology resilience.

[0048] (2) This invention uses the split coefficient index to calculate and evaluate the impact of each key point on the network resilience, and prioritizes the elimination of key points with the greatest impact on the network resilience, so that the limited number of communication relay nodes configured in the network can maximize the role of improving the network resilience.

[0049] (3) By performing algebraic connectivity calculation and analysis on the network topology, this invention selects the communication relay node with the highest redundancy in the current network to carry out the mobile deployment operation to eliminate the key point, which can ensure the rational and efficient use of communication relay node resources. Attached Figure Description

[0050] Figure 1 This is a flowchart of the adaptive dynamic topology damage resistance optimization method of Embodiment 1 of the present invention.

[0051] Figure 2 Network topology connection diagram of Embodiment 2 of the present invention.

[0052] Figure 3The optimized network topology before deployment in Embodiment 2 of the present invention.

[0053] Figure 4 The optimized network topology of Embodiment 2 of the present invention.

[0054] Figure 5 Network topology connection diagram of embodiment 3 of the present invention.

[0055] Figure 6 The optimized network topology before deployment in Embodiment 3 of the present invention.

[0056] Figure 7 The optimized network topology of Embodiment 3 of the present invention. Detailed Implementation

[0057] To provide a clearer understanding of the technical features, objectives, and effects of the present invention, specific embodiments are now described. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention; that is, the described embodiments are only a part of the embodiments of the invention, not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort are within the scope of protection of the present invention.

[0058] Example 1

[0059] like Figure 1 As shown, this embodiment provides an adaptive dynamic topology resilience optimization method for wireless ad hoc networks. First, key nodes in the network are detected, and the impact of each key node on network resilience is calculated and evaluated using a split coefficient index. Then, based on the split coefficient index of the key nodes, key nodes in the network are eliminated sequentially. For each key node in the network, the best communication relay node in the current network is selected and relocated, ensuring that the relocated relay node can effectively connect the connected branches connected to the target key node. By relocating and adjusting the communication relay nodes one by one, key nodes affecting resilience in the network are eliminated sequentially. This method adaptively selects communication relay nodes for relocation and deployment, optimizes and adjusts the network topology in real time, and dynamically enhances the resilience and connectivity of the entire wireless ad hoc network.

[0060] The planned wireless ad hoc network consists of N ordinary nodes and M communication relay nodes. The N ordinary nodes are deployed within a certain area according to application or task requirements, forming a multi-hop wireless ad hoc network. As the nodes move, the network topology continuously and dynamically changes. To achieve real-time adaptive dynamic topology optimization for network resilience, M communication relay nodes (referred to as relay nodes) are planned and deployed in the network. These M relay nodes can move freely within the network as needed for overall topology optimization, adaptively selecting the optimal deployment location to provide communication relay services to the ordinary nodes in the network.

[0061] The communication relay nodes in the network continuously monitor the network topology in real time and perform calculations and analyses based on the monitoring results to implement dynamic topology optimization of the network. For ease of description, the network topology obtained by the communication relay nodes is represented by an undirected graph G = (V, E), where V is the set of nodes in the network and E is the set of connections between nodes in the network.

[0062] Based on the monitoring of network topology by communication relay nodes, the adaptive dynamic topology resilience optimization method proposed in this embodiment consists of four steps: key point detection and evaluation, redundant relay node selection, optimal deployment location determination, and relay node movement and local adjustment. Its main working mechanism is as follows: Figure 1 As shown.

[0063] (1) Joint detection and assessment

[0064] Definition of a key node: In an undirected connected graph G = (V, E), if for a node v ∈ V, after deleting node v and all edges associated with node v from graph G, graph G splits into two or more unconnected subgraphs, then node v is a key node of graph G.

[0065] Detecting key nodes in a network topology graph can be achieved using a naive method of judging each node according to the above definition, or it can be done quickly using the classic Tarjan algorithm proposed in related research. The presence of key nodes in a network is a major reason for its poor resilience. Once a key node fails or is damaged, the network will be split into multiple connected branches, and all communication nodes will no longer form a complete and connected network. Therefore, key nodes are critical nodes affecting network connectivity. Eliminating key nodes in the network through the reasonable deployment of communication relay nodes can effectively improve network resilience.

[0066] Suppose that N is found in the network through key point detection. a If there are 1 key point, then the node set V is used. a The process is represented as |V a |=N aDifferent nodes in a network have varying impacts on network resilience. When the planned number of communication relay nodes is limited, it is necessary to prioritize eliminating nodes with a higher impact on network resilience, based on the severity of their impact.

[0067] We define a splitting coefficient index *e* for each key node. This index expresses the degree to which the network is split after a key node is removed, thus measuring the impact of key nodes on the network's resilience. Suppose that after a key node is removed, the undirected graph G is split into *l* connected components, and the *k*th connected component is denoted as Gk. k The number of nodes in each connected component are N1, N2, ..., N. l The formula for calculating the splitting coefficient index e is as follows:

[0068]

[0069] In the above formula, N(N-1) represents the number of source-destination node communication pairs that an undirected graph G can support, N k (N k -1) represents the connected component G. k The number of source-destination node communication pairs that can be supported. It's easy to see that due to the removal of key nodes, the number of source-destination node communication pairs that the network can support decreases. The split coefficient index e reflects the impact of key node removal on the number of supported communication node pairs. The larger the split coefficient index e, the greater the impact of key nodes on network resilience.

[0070] For each set V a For each keypoint, a splitting coefficient index e is calculated, and the keypoints are sorted according to the magnitude of the splitting coefficient index e. Subsequently, the keypoints will be eliminated in descending order of the splitting coefficient index e.

[0071] (2) Selection of redundant relay nodes

[0072] There are M communication relay nodes in the network that can be used to optimize the network topology, thereby eliminating bottlenecks in the network topology graph. However, at any given moment during network operation, not every communication relay node is available for topology optimization. The removal of some relay nodes can create new bottlenecks in the network, introducing new connectivity problems. Only redundant communication relay nodes can be scheduled for topology optimization deployment. Therefore, it is necessary to select appropriate redundant communication relay nodes to perform the topology optimization task.

[0073] Preferably, the steps for selecting redundant relay nodes are as follows:

[0074] Step S201. Initialize the set

[0075] Step S202. Target each communication relay node Rv in the network one by one. i Delete node Rv i Next, check if any new keypoints have been generated in the network, where i = 1, 2, 3, ..., M; if no new keypoints have been generated, then node Rv is removed. i Add to set V c In the set V c It includes all redundant and movable communication relay nodes;

[0076] Step S203. Select set V c The communication relay node Rv with the least impact on network topology resilience k Priority scheduling is used to perform topology-optimized move deployments; for set V c Each node v in i ∈V c Delete node v i The network topology diagram after that is represented as G. i Calculate each graph G i The algebraic connectivity λ(G) i ), take the maximum value max{λ(G) i ), 1≤i≤|V c The node v corresponding to |} i As Rv k If multiple nodes have the same maximum value, then the node closest to the target location determined by the optimized deployment will be selected.

[0077] Preferably, the method for calculating algebraic connectivity includes:

[0078] If a network has n nodes, A(G) is the adjacency matrix representation of the network topology graph G:

[0079]

[0080] Among them, a ij =1 indicates node v i ,v j There is a link between them; otherwise, a ij =0;

[0081] Let D(G) denote the diagonal matrix formed by the node degrees of each node in graph G:

[0082]

[0083] Among them, node degree

[0084] Let L(G) denote the Laplacian matrix of graph G, and its calculation is expressed as:

[0085] L(G) = D(G) - A(G)

[0086] The eigenvalues ​​of the Laplace matrix L(G) are all non-negative real numbers, with the smallest eigenvalue being 0. Its n eigenvalues, after being sorted, are represented as 0 = λ₁ ≤ λ₂ ≤ ... ≤ λₙ. n The algebraic connectivity λ(G) of graph G is the second smallest eigenvalue of the Laplace matrix L(G), that is:

[0087] λ(G)=λ2

[0088] A larger algebraic connectivity λ(G) indicates better network connectivity and a more robust network topology. The steps described above selected the communication relay node with the highest algebraic connectivity in the network topology after node deletion, signifying that this relay node has the highest redundancy.

[0089] (3) Optimize deployment location determination

[0090] The communication relay node Rv selected in the previous step k To optimize the elimination of the target key v a Mobile deployment in the vicinity enhances connectivity at key nodes. a The connected branches WG k To achieve better connectivity, it is necessary to determine the target location for the mobile deployment of communication relay nodes.

[0091] For target joint v a After removing the p connected branches, the target location for optimizing the deployment of the communication relay node should be a balanced compromise position between the p nearest nodes in these p connected branches, ensuring that the communication relay node can establish a good link connection with all p connected branches. Preferably, in this embodiment, the geometric center of the coordinates of the p nearest nodes is selected as the target location for optimized deployment.

[0092] Assume the key point v a The p connected components formed after deletion are WG1, WG2, ..., WG p Their node sets are represented as V1, V2, ..., V p To find the nearest node to these p connected components and determine its geometric center, this embodiment employs the following steps:

[0093] Step S301. Record the target location for optimized deployment of the communication relay node using R(x,y), and use D... m Record the sum of the squared minimum distances to each connected component, and assign an initial value D. m =+∞;

[0094] Step S302. Sequentially from the connected component WGi Select one node from each of the following groups, and represent them as v1, v2, ..., v p Node v i The corresponding coordinate position is (x i ,y i ), where 1≤i≤p; compute node v i The geometric center position A(x) of the closed geometric shape formed d ,y d ):

[0095]

[0096] Step S303. Calculate the geometric center position A(x) d ,y d ) to each connected branch node v i The sum of the squares of the Euclidean distances is denoted by D:

[0097]

[0098] Step S304. Compare D and D m Calculate the size, take the minimum value, and update D. m Record, i.e., D m =min(D,D) m ); If there is D after the update m =D, then save D. m The target position corresponding to the value is the geometric center position A(x). d ,y d That is, R(x,y)=A(x) d ,y d );

[0099] Step S305. Repeat steps S302 to S304 until all connected components WG have been traversed. i The selection and combination of all nodes in the algorithm are calculated, and the final result R(x,y) is output, which is the target location for the optimized deployment of communication relay nodes.

[0100] The above calculation process is explained using two-dimensional spatial coordinates, but the method is also applicable to three-dimensional spatial coordinates.

[0101] (4) Relay node movement and local adjustment

[0102] After calculating the optimal deployment location R(x,y) for the communication relay node, the selected communication relay node Rv k The deployment will proceed to the target location, with local adjustments made based on the actual connectivity at the target location to ensure effective connectivity across all branches. The steps are as follows:

[0103] Step S401. Communication relay node Rv k Autonomously move to the target location;

[0104] Step S402. If it is not possible to achieve connection with all connected branches WG i If the nodes are connected, then the communication relay node Rv k To optimize the elimination of key object v a The location is locally moved and adjusted until it is connected to all connected branches WG. i All connections of nodes; in the minimum case, relay node Rv k Will move to the joint v a The location is the same as the joint point v. a Similarly, establish connections with all connected branches.

[0105] After eliminating one key node, if there are still relevant nodes to be eliminated in the network, and there are also redundant and movable communication relay nodes, then the above four steps are iteratively executed to continuously optimize the network topology until all key nodes are eliminated.

[0106] Example 2

[0107] This embodiment, based on Embodiment 1, further illustrates the adaptive dynamic topology resilience optimization method with a network example. It should be noted that this embodiment uses a network example as a scenario to specifically illustrate the implementation and operation process of the adaptive dynamic topology resilience optimization method, but the method of this invention is not limited to use in this example network.

[0108] like Figure 2 The diagram illustrates a wireless ad hoc network topology, where dashed lines represent one-hop links for communication between nodes, and numbers indicate node numbers. In this example, seven ordinary nodes form a multi-hop mesh topology.

[0109] against Figure 2 The network shown, through keypoint detection and evaluation, has two keypoints: node 4 and node 5. The split coefficients for keypoints 4 and 5 are denoted as e4 and e5, respectively, and are calculated as e4 = 0.71 and e5 = 0.52. A larger split coefficient indicates a greater impact of keypoint failure on network resilience, and a weaker ability to maintain communication among the remaining nodes. Figure 2 It is clear that a failure at node 4 has a greater impact on the network than a failure at node 5.

[0110] To add two communication relay nodes to the network, such as Figure 3As shown, nodes R1 and R2 are respectively. In the current network, deleting nodes R1 and R2 will not create new nodes, so nodes R1 and R2 can be used for topology optimization relocation deployment during the selection of redundant relay nodes. Through the calculation in step S203, the algebraic connectivity after deleting nodes R1 and R2 are λ(G1) = 0.52 and λ(G2) = 0.44, respectively. Node R1 with the maximum value is prioritized for topology optimization relocation deployment to eliminate the target node 4 with the largest split coefficient. After the operation of optimizing the deployment location, the removal of target node 4 forms two connected branches, and the two nearest nodes in these two connected branches are nodes 3 and 5. Therefore, the target location for the optimized deployment of communication relay node R1 is the geometric center of the line connecting nodes 3 and 5. Finally, the relay node movement and local adjustment operation is performed. Node R1 moves to the target location and, based on the actual connectivity with nodes 3 and 5, makes a local adjustment to the location of node 4.

[0111] After optimizing and eliminating key node 4, key node 5 still needs to be eliminated, along with redundant and movable communication relay node R2. Therefore, dynamic topology resilience optimization is continued, eliminating key node 5 through the relocation and deployment of node R2. The optimized network topology is as follows. Figure 4 As shown.

[0112] Example 3

[0113] This embodiment, based on Embodiment 1, further illustrates the adaptive dynamic topology resilience optimization method with a network example. It should be noted that this embodiment uses a network example as a scenario to specifically illustrate the implementation and operation process of the adaptive dynamic topology resilience optimization method, but the method of this invention is not limited to use in this example network.

[0114] like Figure 5 The diagram shows another wireless ad hoc network topology, in which 13 ordinary nodes form a multi-hop mesh topology.

[0115] against Figure 5 The network shown has three key nodes, identified through key node detection and evaluation: nodes 5, 10, and 12. The splitting coefficients of nodes 5, 10, and 12 are denoted as e5, e2, and e3, respectively. 10 e 12 The calculations are as follows: e5 = 0.77, e 10 =0.50, e 12 =0.29.

[0116] Add three communication relay nodes to the network, such as Figure 6As shown, nodes R1, R2, and R3 are respectively. In the current network, deleting nodes R1, R2, and R3 will not create new nodes. Therefore, nodes R1 / R2 / R3 can all be used for topology optimization mobile deployment during the selection of redundant relay nodes. Through the calculation in step S203, the algebraic connectivity calculations after deleting nodes R1 / R2 / R3 are λ(G1) = 0.177, λ(G2) = 0.204, and λ(G3) = 0.177, respectively. Node R2, corresponding to the maximum value, is prioritized for topology optimization mobile deployment to eliminate the target node 5 with the largest split coefficient. Through the operation of determining the optimized deployment position, after deleting the target node 5, three connected branches are formed. The three nearest nodes in these three connected branches are nodes 3, 7, and 10. Therefore, the target position for the optimized deployment of communication relay node R2 is the geometric center of the shape formed by the line connecting nodes 3, 7, and 10. Finally, the relay node movement and local adjustment operation is performed. Node R2 moves to the target position and, based on the actual connectivity with nodes 3, 7, and 10, makes a local movement and adjustment to the position of joint 5.

[0117] After optimizing and eliminating key node 5, nodes 10 and 12 still need to be eliminated, along with redundant and movable communication relay nodes R1 and R3. Therefore, the dynamic topology resilience optimization operation continues. By moving node R1, target key node 10 is eliminated, and key node 12 is also eliminated simultaneously. The optimized network topology is as follows. Figure 7 As shown.

[0118] It should be noted that, for the sake of simplicity, the foregoing method embodiments are described as a series of actions. However, those skilled in the art should understand that this application is not limited to the described order of actions, as some steps may be performed in other orders or simultaneously according to this application. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are preferred embodiments, and the actions and modules involved are not necessarily essential to this application.

Claims

1. An adaptive dynamic topology invulnerability optimization method for a wireless ad hoc network, characterized in that, The method comprises the following steps: S1. Key node detection and evaluation: detecting key nodes in the network, and calculating and evaluating the influence of each key node on the invulnerability of the network by using a splitting coefficient index, and then preferentially selecting and removing key nodes in the network in order based on the splitting coefficient index of the key nodes; S2. Redundant relay node selection: if the algebraic connectivity of the network topology is the largest after a certain communication relay node is removed, the communication relay node has the highest redundancy, and the communication relay node is selected for topology optimization; S3. Determining the optimal deployment position: determining the target position of the optimal deployment of the communication relay node, so that the communication relay node after moving can effectively connect the connected branches connected by the target key node; S4. Moving and local adjustment of the relay node: moving and deploying the communication relay node to the target position, and performing local adjustment according to the actual connectivity obtained at the target position to effectively connect each connected branch, thereby eliminating the key nodes affecting the invulnerability of the network; The network topology obtained by monitoring the communication relay node is represented as an undirected graph G = ( V , E ) , wherein V is a set of nodes in the network, E is a set of connections between nodes in the network; if for a node v ∈ V , after removing node G and all edges associated with node v from the undirected graph v , the undirected graph G is split into two or more unconnected subgraphs, then node v is a critical node of the undirected graph G ; In step S1, if a node set N a is found to exist in the network by the node detection, then the node set V a is represented as V a N a If the node set is deleted, the undirected graph G is split into l connected branches, the k connected branch is represented as G k , and the number of nodes in each connected branch is N 1, N 2, N l , then the calculation formula of the split coefficient index e is as follows:​ In the above formula, N is the number of ordinary nodes in the network, N N -1) denotes an undirected graph G the source-destination node communication pairs that can be supported, N k N k -1) denotes a connected component G k the source-destination node communication pairs that can be supported, the splitting factor index e The greater the degree of influence of the key node on the network invulnerability.​​ If the joint v a After the deletion p The number of connected branches respectively is WG 1, WG 2, …, WG p The node set is respectively represented as V 1, V 2, …, V p The step S3, i.e. the optimization of the deployment position determination, comprises the following sub-steps: Step S301. Record the target position of the communication relay node optimization deployment, to R x y D m Record the minimum sum of square distances to each connected branch, assign initial value D m =+∞;​​​ Step S302. Sequentially from the connected components WG i Select one node from each of the following, and represent them as follows: v 1, v 2, …, v p ,node v i The corresponding coordinates are ( x i , y i ), where 1≤ i ≤ p ; Computing node v i Geometric center position of the constituted closed geometric shape A ( x d , y d ) Step S303. Calculate the geometric center position A ( x d , y d ) to the square sum of the Euclidean distance of each connected branch node v i , recorded as D : Step S304. Compare D and D m Size, take the minimum value, and update D m Record, that is, there D m = min( D , D m ); when updated, if there D m = D , save D m The target position corresponding to the value is the geometric center position A ( x d , y d ), that is, there R ( x , y )= A ( x d , y d ) Step S305. Repeat the execution of steps S302-S304 until all the connected branches are traversed WG i all the node selection combinations and calculations, and output the final result R ( x , y ), which is the target position for the optimized deployment of the communication relay node.

2. The method of claim 1, wherein, one by one for the set V a Split coefficient index for each joint in e Calculate and based on the splitting coefficient index e The key points are sorted by size, and subsequent sorting will be based on the splitting coefficient index. e The joints are eliminated sequentially from largest to smallest.

3. The method of claim 1, wherein, Step S2, i.e. redundant relay node selection, comprises the following sub-steps: Step S201. Initialize the set V c = ; Step S202. For each communication relay node in the network, one by one Rv i , delete the node Rv i After that, detect whether a new relevant node is generated in the network, wherein i =1, 2, 3, …, M , M is the number of communication relay nodes; if a new relevant node is not generated, add the node Rv i to the set V c , and the set V c contains all the redundant movable communication relay nodes; Step S203. Selecting the set V c The communication relay node with the least impact on the network topology Rv k Prioritizing the scheduling for performing the topology-optimized mobile deployment; for each node in the set V c The node v i ∈ V c , deleting the node v i The network topology graph after the node is represented as G i Calculating the algebraic connectivity of each graph G i The maximum value of the algebraic connectivity The node corresponding to the maximum value v i As Rv k If there are multiple nodes corresponding to the same maximum value, the node closest to the target position determined according to the optimized deployment is selected.​ 4. The method of claim 3, wherein, In step S203, the calculation method of the algebraic connectivity comprises: If a network has n nodes, A ( G ) is an adjacency matrix representation of the network topology graph G ​ wherein, a ij = 1 indicates that there is a link between nodes v i , v j = 0; otherwise, a ij = 0; With D ( G ) represents a graph G Diagonal matrix of the node degrees of the nodes wherein the node degree ; The Laplacian matrix of a graph L ( G ) is represented by G The Laplacian matrix of a graph whose computation is represented by laplacian matrix L ( G ) are non-negative real numbers, the smallest eigenvalue is zero, and the n eigenvalues, ordered, are denoted by ; the algebraic connectivity G of the graph is the second smallest eigenvalue of the laplacian matrix L ( G ), i.e. there is: algebraic connectivity The larger, the better the connectivity of the network, and the network topology has better invulnerability.

5. The method of claim 1, wherein, In step S302, the node v i The geometric center position of the closed geometric shape thus formed A ( x d , y d ) is: 。 6. The method of claim 1, wherein, Step S4, i.e. moving and local adjustment of the relay node, comprises the following sub-steps: Step S401. A communication relay node Rv k Autonomously move to the target location; Step S402. If the connection to all the communicating branches of the node is not achieved WG i The communication relay node is moved to the location of the node Rv k The node object is optimized for elimination v a The location of the node is locally adjusted until the connection to all the communicating branches of the node is achieved WG i The connection to all the communicating branches of the node is achieved; at a minimum, the relay node Rv k Will be moved to the location of the node v a The connection to all the communicating branches of the node is achieved; at a minimum, the relay node v a Will be moved to the location of the node 7. The method of claim 1-6, wherein, After the elimination of one key node is completed, if there are still key nodes to be eliminated in the network, and there are also redundant communication relay nodes that can be moved, steps S1-S4 are iteratively executed to continuously optimize the network topology until all key nodes are eliminated.

Images