A method for adaptively adjusting topology of wireless sensor networks

Abstract

The present invention relates to the field of wireless communication technology, and specifically discloses a method for adaptively adjusting the topology of a wireless sensor network, which continuously detects whether a node is interfered with by a signal, divides the nodes that are not interfered with by the signal into clusters, and selects cluster heads, thereby realizing communication of a hierarchical topology and adaptive adjustment. Specifically, when selecting a cluster head, the present invention selects by establishing a topological relationship description graph with the minimum transmission cost between clusters. The connection relationship of the vertices in the topological relationship description graph characterizes the communication connection relationship between each cluster. At this time, based on these connection relationships, multiple cluster heads can be determined in each cluster. On the one hand, multiple cluster heads improve the fault tolerance rate of the entire network and enhance its robustness. On the other hand, multiple cluster heads also reduce the communication pressure of each cluster head node, reduce the load level of the network, and reduce the energy consumption of the cluster head, which has good practicality.

Description

Technical Field

[0001] The present invention relates to the technical field of wireless communications, and in particular to a method for adaptively adjusting topology of a wireless sensor network. Background Art

[0002] A wireless sensor network is a network consisting of a group of wireless sensor nodes that can connect to each other through wireless communication and collect, process and transmit data in the environment. This network is usually used to monitor and control various environments, such as industrial production, urban infrastructure, military applications, etc. Sensor nodes can be various types of sensors, such as temperature sensors, humidity sensors, light sensors, etc. They transmit the collected data to other nodes or data centers in the network through wireless communication for further analysis and application.

[0003] Hierarchical topology control is a typical topology control method for wireless sensor networks. The LEACH (Low Energy Adaptive Clustering Hierarchy) protocol is one of them, which proposes the idea of ​​clustering. In the LEACH protocol, the nodes in the network are divided into cluster structures. Each cluster is managed by a cluster head node, which is responsible for transmitting the data in the cluster to other cluster heads or directly to the aggregation node. The data sensed by the ordinary node will be transmitted to the cluster head node of the cluster to which it belongs. The cluster head nodes form a backbone network that receives, aggregates and forwards data. In this way, the LEACH protocol significantly reduces the number of transmissions, thereby reducing the overall energy consumption of the WSN network and extending the network's survival time.

[0004] Obviously, the selection of cluster heads is the key to determine the effect of hierarchical topology control. However, in the current hierarchical topology control method, usually only one cluster head is selected for each cluster, which makes the stability of the entire network poor. If a cluster head node cannot work normally due to signal interference, all nodes in the entire cluster will be paralyzed. Therefore, people need a wireless sensor network topology adaptive adjustment method that can determine multiple cluster heads. Summary of the invention

[0005] The purpose of the present invention is to provide a method for adaptively adjusting the topology of a wireless sensor network to solve the following technical problems:

[0006] The existing hierarchical topology control methods cannot determine multiple cluster heads and have poor stability.

[0007] The purpose of the present invention can be achieved through the following technical solutions:

[0008] A method for adaptively adjusting topology of a wireless sensor network comprises the following steps:

[0009] Acquire multiple nodes in the target wireless sensor network that are not interfered by signals as target nodes, and divide the multiple target nodes into multiple clusters;

[0010] Obtain the central node of each cluster, and based on the positional relationship of multiple central nodes and the number of target nodes in each cluster, establish a topological relationship description diagram with the minimum transmission cost between clusters;

[0011] Determine the communication connection relationship between each cluster according to the connection relationship of each vertex in the topological relationship description graph, and determine multiple cluster heads of each cluster according to the position relationship between multiple clusters;

[0012] The cluster heads are used to communicate between two clusters based on hierarchical topology, and each node is monitored for signal interference.

[0013] If the proportion of target nodes affected by signal interference exceeds a first preset threshold, the steps of acquiring multiple nodes in the target wireless sensor network that are not affected by signal interference and dividing the multiple nodes into multiple clusters are performed again.

[0014] As a further solution of the present invention: the step of obtaining a plurality of nodes in the target wireless sensor network that are not interfered by signals as target nodes and dividing the plurality of target nodes into a plurality of clusters includes:

[0015] Acquire multiple nodes in the target wireless sensor network that are not interfered by signals as target nodes, and acquire the coverage area of ​​the target wireless sensor network and the average communication range of each target node;

[0016] According to the coverage area of ​​the target wireless sensor network and the proportional relationship between the average communication ranges of multiple target nodes, the neighborhood radius and the minimum number of data points are determined;

[0017] According to the neighborhood radius and the minimum number of data points, and the distance between target nodes, multiple target nodes are clustered using DBSCAN to obtain multiple clusters and noise points.

[0018] The noise points are assigned to the clusters that are closest to them and can be communicated with, and multiple clusters consisting of target nodes are obtained.

[0019] As a further solution of the present invention: determining the neighborhood radius and the minimum number of data points according to the proportional relationship between the coverage area of ​​the target wireless sensor network and the average communication range of multiple target nodes includes:

[0020] Determine a neighborhood radius according to an average communication range of the plurality of target nodes, wherein the neighborhood radius is smaller than the average communication range of the plurality of target nodes;

[0021] The minimum number of data points is determined by the following formula:

[0022]

[0023] Among them, Nmin is the minimum number of data points, N is the preset lower limit, A is the coverage area of ​​the target wireless sensor network, r is the average communication range of multiple target nodes, and a and b are different weight parameters.

[0024] As a further solution of the present invention: the central node of each cluster is obtained, and based on the position relationship of multiple central nodes and the number of target nodes of each cluster, a topological relationship description diagram with the minimum transmission cost between clusters is established, including:

[0025] Get the central node of each cluster and its corresponding position relationship;

[0026] According to the position relationship of the central nodes, a cluster topology is established. Each vertex in the cluster topology corresponds to a central node. The central nodes corresponding to the two vertices connected by each edge in the cluster topology can communicate with each other. The weight of each edge in the cluster topology is the distance between the central nodes corresponding to the two vertices connected by the edge.

[0027] Find a minimum spanning tree of the cluster topology graph as a topological relationship description graph.

[0028] As a further solution of the present invention: the method of finding a minimum spanning tree of the cluster topology graph as a topology relationship description graph includes:

[0029] Randomly select a vertex in the cluster topology as the initial vertex and add it to the minimum spanning tree;

[0030] According to the weight of each vertex and edge in the cluster topology, we continuously search for the edge that represents the minimum transmission cost, is connected to the current minimum spanning tree, and does not form a loop, and the vertices connected by the edge, and add them to the minimum spanning tree until all vertices are added to the minimum spanning tree;

[0031] The final minimum spanning tree is used as the topological relationship description graph.

[0032] As a further solution of the present invention: in the cluster topology graph, each vertex also corresponds to a weight, and the weight of the vertex is the number of target nodes in the cluster where the central node corresponding to the vertex is located; the transmission cost represented by each edge in the cluster topology graph is calculated by the following formula:

[0033] COST=c×∑Wvertex+d×Wedge;

[0034] Among them, COST is the transmission cost of an edge, Wvertex is the weight of a vertex connected by the edge, Wedge is the weight of the edge, and c and d are different weight coefficients respectively.

[0035] As a further solution of the present invention: the communication connection relationship between each cluster is determined according to the connection relationship of each vertex in the topological relationship description graph, and multiple cluster heads of each cluster are determined according to the position relationship between multiple clusters, including:

[0036] Find two clusters corresponding to two adjacent vertices in the topological relationship description graph as two target clusters with communication connection relationship;

[0037] According to the positions of the target nodes in the two target clusters, two target nodes with the closest distances are selected from the two target clusters and serve as cluster heads of the two target clusters respectively.

[0038] As a further solution of the present invention: also include:

[0039] If the proportion of target nodes affected by signal interference does not exceed the preset threshold, the target nodes affected by signal interference are put into sleep mode.

[0040] Determine whether the dormant node is the cluster head of its cluster;

[0041] If the dormant node is the cluster head of its cluster, then an adjacent node of the dormant node is selected as the new cluster head.

[0042] As a further solution of the present invention: also include:

[0043] At preset intervals, the dormant nodes are reawakened and re-detected to see if the reawakened nodes are still subject to signal interference;

[0044] When the number of nodes that are reawakened and not interfered by signals exceeds a second preset threshold, the steps of acquiring multiple nodes that are not interfered by signals in the target wireless sensor network and dividing the multiple nodes into multiple clusters are performed again.

[0045] Beneficial effects of the present invention:

[0046] The present invention provides a method for adaptively adjusting the topology of a wireless sensor network, which continuously detects whether a node is interfered with by a signal, divides the nodes that are not interfered with by the signal into clusters, and selects cluster heads, thereby realizing communication of a hierarchical topology and adaptive adjustment. Specifically, when selecting a cluster head, the present invention selects by establishing a topological relationship description graph with the minimum transmission cost between clusters. The connection relationship of the vertices in the topological relationship description graph represents the communication connection relationship between each cluster. At this time, according to these connection relationships, multiple cluster heads can be determined in each cluster. On the one hand, multiple cluster heads improve the fault tolerance rate of the entire network and enhance its robustness. On the other hand, multiple cluster heads also reduce the communication pressure of each cluster head node, reduce the load level of the network, and reduce the energy consumption of the cluster head, which has good practicality. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] The present invention will be further described below in conjunction with the accompanying drawings.

[0048] Figure 1 It is a method flow chart of the method for adaptively adjusting the topology of a wireless sensor network of the present invention;

[0049] Figure 2 yes Figure 1 The method flow chart of step S101 in FIG.

[0050] Figure 3 yes Figure 1 The method flow chart of step S102 in FIG.

[0051] Figure 4 It is a schematic diagram of a cluster head obtained by the prior art;

[0052] Figure 5 A schematic diagram of a cluster head obtained by the method of the present invention is DETAILED DESCRIPTION

[0053] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0054] Referring to Figure 1, the present invention is a method for adaptively adjusting the topology of a wireless sensor network, comprising the following steps:

[0055] S101, acquiring multiple nodes in a target wireless sensor network that are not interfered by signals as target nodes, and dividing the multiple target nodes into multiple clusters;

[0056] S102, obtaining the central node of each cluster, and establishing a topological relationship description diagram with the minimum transmission cost between clusters based on the position relationship of multiple central nodes and the number of target nodes of each cluster;

[0057] S103, determining the communication connection relationship between each cluster according to the connection relationship between each vertex in the topological relationship description graph, and determining multiple cluster heads of each cluster according to the position relationship between the multiple clusters;

[0058] S104, performing communication between two clusters based on the hierarchical topology based on the cluster head, and monitoring whether each node is interfered with by the signal;

[0059] S105: If the proportion of target nodes affected by signal interference exceeds a first preset threshold, re-execute the step of acquiring multiple nodes in the target wireless sensor network that are not affected by signal interference and dividing the multiple nodes into multiple clusters.

[0060] The present invention provides a method for adaptively adjusting the topology of a wireless sensor network, which continuously detects whether a node is interfered with by a signal, divides the nodes that are not interfered with by the signal into clusters, and selects cluster heads, thereby realizing communication of a hierarchical topology and adaptive adjustment. Specifically, when selecting a cluster head, the present invention selects by establishing a topological relationship description graph with the minimum transmission cost between clusters. The connection relationship of the vertices in the topological relationship description graph represents the communication connection relationship between each cluster. At this time, according to these connection relationships, multiple cluster heads can be determined in each cluster. On the one hand, multiple cluster heads improve the fault tolerance rate of the entire network and enhance its robustness. On the other hand, multiple cluster heads also reduce the communication pressure of each cluster head node, reduce the load level of the network, and reduce the energy consumption of the cluster head, which has good practicality.

[0061] In addition, it should be noted that how to detect whether a signal is interfered is an existing technology that can be understood by those skilled in the art, so it will not be explained in detail in this article.

[0062] Further, combined with Figure 2 As shown, in a preferred embodiment, the above step S101, obtaining multiple nodes in the target wireless sensor network that are not interfered by the signal as target nodes, and dividing the multiple target nodes into multiple clusters, specifically includes:

[0063] S201, obtaining multiple nodes in the target wireless sensor network that are not interfered by signals as target nodes, and obtaining the coverage area of ​​the target wireless sensor network and the average communication range of each target node;

[0064] S202, determining a neighborhood radius and a minimum number of data points according to a proportional relationship between a coverage area of ​​a target wireless sensor network and an average communication range of a plurality of target nodes;

[0065] S203, performing DBSCAN clustering on multiple target nodes according to the neighborhood radius and the minimum number of data points and the distances between the target nodes to obtain multiple clusters and noise points;

[0066] S204: Allocate the noise point to the cluster that is closest to it and can be communicated with, to obtain multiple clusters consisting of target nodes.

[0067] The above process uses the DBSCAN algorithm to cluster similar nodes together according to distance to form a cluster. DBSCAN (Density-Based Spatial Clustering of Applications with Noise) is a density-based clustering algorithm that aims to divide data points into dense areas (clusters) and sparse areas (noise). Unlike other clustering algorithms, DBSCAN does not need to pre-specify the number of clusters, but automatically discovers the shape and number of clusters based on the density of data points. DBSCAN can handle clusters of any shape, which makes it more flexible in practical applications and is not limited by the shape of clusters. DBSCAN does not need to pre-specify the number of clusters. Compared with some algorithms that need to determine the number of clusters in advance, such as K-means, DBSCAN is more flexible. Although DBSCAN needs to adjust two parameters (neighborhood radius and minimum number of data points), it has lower requirements for parameter selection than other algorithms. It can be seen that DBSCAN is particularly suitable for the application scenario of the present invention.

[0068] Different from the traditional DBSCAN algorithm, the present invention further classifies the nodes represented by some noise points into the clusters they can reach, so as to maximize the application of sensor nodes and avoid the appearance of idle nodes. It is understandable that in practice, the wireless sensor network can also be divided into shapes such as equilateral triangles, pentagons, and regular hexagons to obtain multiple clusters.

[0069] The two key parameters in the DBSCAN (Density-Based Spatial Clustering of Applications with Noise) algorithm are ε (Eps) and MinPts. Among them, ε represents the neighborhood radius, that is, the radius used to determine the ε-neighborhood of a data point. This means that all data points within a range of no more than ε from a data point are considered to be neighbors of the data point. MinPts represents the minimum number of data points in the ε-neighborhood, that is, how many data points are required in the ε-neighborhood of a data point to mark the point as a core object. The setting of these two parameters is crucial to the clustering results of the DBSCAN algorithm. Appropriate ε and MinPts values ​​can affect the accuracy and performance of clustering. Too small ε may cause too many noise points to be classified as a cluster, while too large ε may cause data points to be over-diffused and unable to form an obvious cluster structure; at the same time, appropriate selection of MinPts can help filter out noise points and ensure that only areas with sufficient density are considered clusters. Therefore, by adjusting and optimizing the values ​​of ε and MinPts, we can better adapt to the characteristics of different data sets and obtain more accurate and stable clustering results.

[0070] The present invention provides a preferred determination method for this. Specifically, in a preferred embodiment, the above step S202, according to the coverage area of ​​the target wireless sensor network and the proportional relationship between the average communication ranges of multiple target nodes, determines the neighborhood radius and the minimum number of data points, specifically including:

[0071] Determine a neighborhood radius according to an average communication range of the plurality of target nodes, wherein the neighborhood radius is smaller than the average communication range of the plurality of target nodes;

[0072] The minimum number of data points is determined by the following formula:

[0073]

[0074] Among them, Nmin is the minimum number of data points, N is the preset lower limit, A is the coverage area of ​​the target wireless sensor network, r is the average communication range of multiple target nodes, and a and b are different weight parameters.

[0075] The significance of the above formula is that an optimal method is used to determine the neighborhood radius and the minimum number of data points, that is, the neighborhood radius is determined according to the average communication range to ensure that the nodes in a cluster can communicate normally. In practice, the neighborhood radius is usually 0.2-0.3 times the average communication range, so as to ensure that the communication between clusters is reachable. The minimum number of data points is then determined by the proportional relationship between the coverage area of ​​the target infinite sensor network and the average communication range of multiple target nodes. The larger the coverage area of ​​the target infinite sensor network, the more clusters can be divided, and the larger the average communication range of multiple target nodes, the stronger the communication capability of each node. At this time, the number of divided clusters can be appropriately reduced to reduce the complexity of the network topology.

[0076] Further, combined with Figure 3 As shown, in a preferred embodiment, the above step S102, obtaining the central node of each cluster, and establishing a topological relationship description diagram with the minimum transmission cost between clusters based on the position relationship of multiple central nodes and the number of target nodes of each cluster, specifically includes:

[0077] S301, obtaining the central node of each cluster and its corresponding position relationship;

[0078] S302, based on the positional relationship of the central nodes, a cluster topology is established, each vertex in the cluster topology corresponds to a central node, the central nodes corresponding to the two vertices connected by each edge in the cluster topology can communicate with each other, and the weight of each edge in the cluster topology is the distance between the central nodes corresponding to the two vertices connected by the edge;

[0079] S303, finding a minimum spanning tree of the cluster topology graph as a topology relationship description graph.

[0080] In the above process, the cluster topology diagram represents the communication reachable relationship between all central nodes, that is, the central nodes corresponding to the two connected vertices can realize communication, and the distance between the vertices in the cluster topology diagram represents the distance between the two central nodes, and can also reflect the communication transmission cost between the two nodes. The minimum spanning tree refers to a tree formed by selecting some edges in a given weighted connected graph (i.e., the cluster topology diagram in this embodiment), so that the tree contains all the nodes in the graph and the total weight is the smallest. It can be understood that in this embodiment, the minimum spanning tree refers to a topological connection method in the cluster spectrum diagram that represents the minimum communication transmission cost. The cluster head determined by the topological relationship description diagram can reduce the energy consumption of the entire network.

[0081] Further, in a preferred embodiment, the above step S303, finding a minimum spanning tree of the cluster topology graph as a topology relationship description graph, specifically includes:

[0082] Randomly select a vertex in the cluster topology as the initial vertex and add it to the minimum spanning tree;

[0083] According to the weight of each vertex and edge in the cluster topology, we continuously search for the edge that represents the minimum transmission cost, is connected to the current minimum spanning tree, and does not form a loop, and the vertices connected by the edge, and add them to the minimum spanning tree until all vertices are added to the minimum spanning tree;

[0084] The final minimum spanning tree is used as the topological relationship description graph.

[0085] The above process uses the improved prim algorithm to select the minimum spanning tree. The prim algorithm is a greedy algorithm for constructing the minimum spanning tree. It starts from a vertex in the graph and gradually increases the edges and vertices until all the vertices are included. At each step of the algorithm, the edge with the smallest weight connecting the selected vertices and the unselected vertices is selected. A major advantage of the prim algorithm is its simplicity and efficiency, especially when dealing with dense graphs. The algorithm can be implemented by a priority queue (such as a binary heap) to further optimize the edge selection process, thereby achieving a faster running time. When implementing, the prim algorithm only needs to maintain a priority queue to store all the edges that span the constructed part and the unconstructed part, extract the minimum edge from the queue each time, and update the data structure. This method makes the algorithm very suitable for the minimum spanning tree problem of dense graphs. In addition, the implementation of the prim algorithm can easily adapt to changes in the graph, such as adding or deleting nodes and edges, making it flexible and applicable.

[0086] Compared with the common Prim algorithm, in this embodiment, when selecting the edge to be added to the minimum spanning tree, the edge representing the minimum transmission cost is selected, so that a topological relationship description graph with the minimum transmission cost can be obtained.

[0087] It is understandable that in practice, the weight of the edge, that is, the distance, can be directly used as the transmission cost, and the present invention also provides a preferred solution, specifically:

[0088] In the cluster topology, each vertex also corresponds to a weight, and the weight of the vertex is the number of target nodes in the cluster where the central node corresponding to the vertex is located; the transmission cost represented by each edge in the cluster topology is calculated by the following formula:

[0089] COST=c×∑Wvertex+d×Wedge;

[0090] Among them, COST is the transmission cost of an edge, Wvertex is the weight of a vertex connected by the edge, Wedge is the weight of the edge, and c and d are different weight coefficients respectively.

[0091] The significance of the above formula is that when selecting the edge with the smallest transmission cost, not only the weight of the edge itself is considered (that is, the larger the weight of the edge, the farther the distance between the two nodes, and the higher the cost), but also the weight of the node is considered. That is, the larger the node weight, the more nodes there are in the cluster, and the more difficult it will be to forward information within the cluster, and the greater the cost.

[0092] Further, in a preferred embodiment, the above step S103, determining the communication connection relationship between each cluster according to the connection relationship between each vertex in the topological relationship description graph, and determining multiple cluster heads of each cluster according to the position relationship between the multiple clusters, includes:

[0093] Find two clusters corresponding to two adjacent vertices in the topological relationship description graph as two target clusters with communication connection relationship;

[0094] According to the positions of the target nodes in the two target clusters, two target nodes with the closest distances are selected from the two target clusters and serve as cluster heads of the two target clusters respectively.

[0095] In the process of selecting two target nodes with the closest distance from each other from the two target clusters, multiple pairs of nodes can be selected from the two clusters to achieve the effect of improving stability.

[0096] Figure 4 Schematic diagram of a cluster head obtained by the prior art, wherein nodes A, B, C and D are the cluster heads of cluster 1, cluster 2, cluster 3 and cluster 4 respectively. It can be seen that once a problem occurs with node A, cluster 2, cluster 3 and cluster 4 cannot communicate normally.

[0097] Figure 5 The figure is a schematic diagram of a cluster head obtained by the method of the present invention, wherein nodes E, F and G respectively realize the communication between cluster 1 and cluster 2, cluster 3 and cluster 4 through nodes X, Y and Z. It can be seen that even if two of nodes E, F and G fail, for example, node E and node G fail, the network can still ensure normal communication between cluster 1 and cluster 3.

[0098] Furthermore, in a preferred embodiment, the wireless sensor network topology adaptive adjustment method further includes:

[0099] If the proportion of target nodes affected by signal interference does not exceed the preset threshold, the target nodes affected by signal interference are put into sleep mode.

[0100] Determine whether the dormant node is the cluster head of its cluster;

[0101] If the dormant node is the cluster head of its cluster, then an adjacent node of the dormant node is selected as the new cluster head.

[0102] The above process is a processing method of the present invention when the cluster head node is disturbed, which can improve the robustness of the present invention.

[0103] Furthermore, in a preferred embodiment, the wireless sensor network topology adaptive adjustment method further includes:

[0104] At preset intervals, the dormant nodes are reawakened and re-detected to see if the reawakened nodes are still subject to signal interference;

[0105] When the number of nodes that are reawakened and not interfered by signals exceeds a second preset threshold, the steps of acquiring multiple nodes that are not interfered by signals in the target wireless sensor network and dividing the multiple nodes into multiple clusters are performed again.

[0106] The above process is a processing method of the present invention when there are too many dormant nodes, which can improve the dynamics and reliability of the entire network system.

[0107] The present invention provides a method for adaptively adjusting the topology of a wireless sensor network, which continuously detects whether a node is interfered with by a signal, divides the nodes that are not interfered with by the signal into clusters, and selects cluster heads, thereby realizing communication of a hierarchical topology and adaptive adjustment. Specifically, when selecting a cluster head, the present invention selects by establishing a topological relationship description graph with the minimum transmission cost between clusters. The connection relationship of the vertices in the topological relationship description graph represents the communication connection relationship between each cluster. At this time, according to these connection relationships, multiple cluster heads can be determined in each cluster. On the one hand, multiple cluster heads improve the fault tolerance rate of the entire network and enhance its robustness. On the other hand, multiple cluster heads also reduce the communication pressure of each cluster head node, reduce the load level of the network, and reduce the energy consumption of the cluster head, which has good practicality.

[0108] The above is a detailed description of an embodiment of the present invention, but the content is only a preferred embodiment of the present invention and cannot be considered to limit the scope of implementation of the present invention. All equivalent changes and improvements made within the scope of the present invention should still fall within the scope of the patent coverage of the present invention.

Claims

1. A method for adaptively adjusting topology of a wireless sensor network, characterized in that: The following steps are involved: Acquire multiple nodes in the target wireless sensor network that are not interfered by signals as target nodes, and divide the multiple target nodes into multiple clusters; Obtain the central node of each cluster, and based on the positional relationship of multiple central nodes and the number of target nodes in each cluster, establish a topological relationship description diagram with the minimum transmission cost between clusters; Determine the communication connection relationship between each cluster according to the connection relationship of each vertex in the topological relationship description graph, and determine multiple cluster heads of each cluster according to the position relationship between multiple clusters; The cluster heads are used to communicate between two clusters based on hierarchical topology, and each node is monitored for signal interference. If the proportion of target nodes affected by signal interference exceeds a first preset threshold, re-performing the step of acquiring multiple nodes in the target wireless sensor network that are not affected by signal interference and dividing the multiple nodes into multiple clusters; The central node of each cluster is obtained, and based on the positional relationship of multiple central nodes and the number of target nodes of each cluster, a topological relationship description diagram with the minimum transmission cost between clusters is established, including: Get the central node of each cluster and its corresponding position relationship; According to the position relationship of the central nodes, a cluster topology is established. Each vertex in the cluster topology corresponds to a central node. The central nodes corresponding to the two vertices connected by each edge in the cluster topology can communicate with each other. The weight of each edge in the cluster topology is the distance between the central nodes corresponding to the two vertices connected by the edge. Find a minimum spanning tree of the cluster topology graph as a topological relationship description graph; The method of determining the communication connection relationship between each cluster according to the connection relationship between each vertex in the topological relationship description graph, and determining multiple cluster heads of each cluster according to the position relationship between multiple clusters, includes: Find two clusters corresponding to two adjacent vertices in the topological relationship description graph as two target clusters with communication connection relationship; According to the positions of the target nodes in the two target clusters, two target nodes with the closest distances are selected from the two target clusters and serve as cluster heads of the two target clusters respectively.

2. The method for adaptively adjusting the topology of a wireless sensor network according to claim 1, characterized in that: The step of acquiring multiple nodes in the target wireless sensor network that are not interfered by signals as target nodes and dividing the multiple target nodes into multiple clusters includes: Acquire multiple nodes in the target wireless sensor network that are not interfered by signals as target nodes, and acquire the coverage area of ​​the target wireless sensor network and the average communication range of each target node; According to the coverage area of ​​the target wireless sensor network and the proportional relationship between the average communication ranges of multiple target nodes, the neighborhood radius and the minimum number of data points are determined; According to the neighborhood radius and the minimum number of data points, and the distance between target nodes, multiple target nodes are clustered using DBSCAN to obtain multiple clusters and noise points. The noise points are assigned to the clusters that are closest to them and can be communicated with, and multiple clusters consisting of target nodes are obtained.

3. The method for adaptively adjusting the topology of a wireless sensor network according to claim 2, characterized in that: The determining of the neighborhood radius and the minimum number of data points according to the coverage area of ​​the target wireless sensor network and the proportional relationship of the average communication ranges of the plurality of target nodes comprises: Determine a neighborhood radius according to an average communication range of the plurality of target nodes, wherein the neighborhood radius is smaller than the average communication range of the plurality of target nodes; The minimum number of data points is determined by the following formula: ； Among them, Nmin is the minimum number of data points, N is the preset lower limit, A is the coverage area of ​​the target wireless sensor network, r is the average communication range of multiple target nodes, and a and b are different weight parameters.

4. The method for adaptively adjusting the topology of a wireless sensor network according to claim 1, characterized in that: The step of searching for a minimum spanning tree of the cluster topology graph as a topology relationship description graph includes: Randomly select a vertex in the cluster topology as the initial vertex and add it to the minimum spanning tree; According to the weight of each vertex and edge in the cluster topology, we continuously search for the edge that represents the minimum transmission cost, is connected to the current minimum spanning tree, and does not form a loop, and the vertices connected by the edge, and add them to the minimum spanning tree until all vertices are added to the minimum spanning tree; The final minimum spanning tree is used as the topological relationship description graph.

5. The method for adaptively adjusting the topology of a wireless sensor network according to claim 4, characterized in that: In the cluster topology, each vertex also corresponds to a weight, and the weight of the vertex is the number of target nodes in the cluster where the central node corresponding to the vertex is located; the transmission cost represented by each edge in the cluster topology is calculated by the following formula: ； Among them, COST is the transmission cost of an edge, Wvertex is the weight of a vertex connected by the edge, Wedge is the weight of the edge, and c and d are different weight coefficients respectively.

6. The method for adaptively adjusting the topology of a wireless sensor network according to claim 1, characterized in that: Also includes: If the proportion of target nodes affected by signal interference does not exceed the preset threshold, the target nodes affected by signal interference are put into sleep mode. Determine whether the dormant node is the cluster head of its cluster; If the dormant node is the cluster head of its cluster, then an adjacent node of the dormant node is selected as the new cluster head.

Images