Unmanned aerial vehicle ad hoc network clustering method and system based on adaptive node degree difference

Abstract

The invention relates to an unmanned aerial vehicle ad hoc network clustering method and system based on adaptive node degree difference. According to the method, the multi-channel characteristic of the large-scale unmanned aerial vehicle ad hoc network is considered, the node similarity introducing the multi-channel constraint condition and the adaptive node degree difference is defined for clustering, and the uniformity of the node connection condition can be reflected through the adaptive node degree difference during clustering, so that the clustering result is more balanced, and the clustering efficiency is improved. The problem of excessive competition or waste of resources caused by excessive maximum clusters or minimum clusters is avoided; furthermore, the connection edges among the nodes in the unmanned aerial vehicle ad hoc network after clustering are circularly processed, the modularity increment brought by cluster merging is dynamically calculated, and the optimal node combination is selected for cluster merging, so that the community structure and the overall performance of the unmanned aerial vehicle networking can be continuously optimized, the cluster scale deviation is reduced, and the optimal clustering result is realized.

Description

Technical Field

[0001] The present application relates to the technical field of drone self-organizing network clustering, and in particular to a drone self-organizing network clustering method and system based on adaptive node degree difference. Background Art

[0002] As the scale of drone swarms continues to expand, the demand for information transmission between drone nodes has increased dramatically, and drone network management faces challenges such as complex and cumbersome routing, high maintenance overhead, and large transmission delays. In addition, under the complex conditions of communication resource shortage and dynamic environmental confrontation, drone self-organizing networks need to perceive external environmental changes in a timely manner and make optimal decisions to achieve efficient sharing of communication resources. Therefore, the contradiction between the continuous expansion of drone self-organizing network communication scale and limited resources has become increasingly prominent.

[0003] The clustering algorithm realizes hierarchical management by dividing the UAV self-organizing network into multiple smaller cluster units. It has the advantages of efficient network management, reusable communication resources, and scalable network scale. It is more suitable for large-scale UAV self-organizing network scenarios, such as performing large-scale UAV formation tasks. Traditional clustering algorithms include the highest node degree clustering, minimum ID clustering, minimum mobility clustering, and weighted clustering algorithms. These algorithms determine the cluster head by node attributes, aiming to build a reasonable network structure and improve network performance. However, these algorithms consider clustering algorithms in an ideal single-channel environment. But in actual applications, due to the complex and changeable spectrum environment and the significant differences in channel availability between nodes, the applicability of traditional single-channel clustering algorithms is limited, and it is difficult to fully meet the needs of the current complex and changeable UAV self-organizing network environment. Summary of the invention

[0004] Based on this, it is necessary to provide a UAV self-organizing network clustering method and system based on adaptive node degree difference to address the above technical problems, aiming to dynamically organize large-scale UAV nodes to achieve efficient communication and resource management in a complex environment with limited spectrum resources and differentiated available channels of nodes.

[0005] A method for clustering unmanned aerial vehicle self-organizing networks based on adaptive node degree difference, the method comprising: According to the available channel set and geographic location information perceived by each node in the UAV self-organizing network, the network topology of the UAV self-organizing network is constructed and the connection edges between each node are obtained; Initialize the modularity and clustering structure of the UAV self-organizing network, calculate the adaptive node degree difference of each node, and calculate the node similarity of introducing multi-channel constraints and adaptive node degree difference, perform clustering according to node similarity, and update the network topology of the UAV self-organizing network; The connection edges between nodes in the UAV self-organizing network are processed cyclically. The number of connection edges is reduced by 1 in each cycle until the number of connection edges is reduced to 0. In each cycle, the modularity increment brought by merging the two clusters corresponding to each connection edge is calculated and recorded. The connection edge with the largest modularity increment and a common channel after merging the clusters in the current cycle is recorded. Add the maximum modularity increment in the current cycle to the modularity of the UAV ad hoc network, and merge the two clusters corresponding to the maximum modularity increment; When the number of connected edges is reduced to 0, the optimal clustering result of the UAV self-organizing network is output.

[0006] In one embodiment, the network topology of the drone ad hoc network includes nodes, the spectrum is divided into Non-overlapping channels, each node is equipped with a half-duplex transceiver for sensing channel availability. Any two nodes can communicate directly if they are within the communication range of each other and have the same available channel. The communication distance between nodes is limited by the maximum communication distance. .

[0007] In one embodiment, the modularity and clustering structure of the drone ad hoc network are initialized, and the adaptive node degree difference of each node is calculated, including: Set the modularity of the drone ad hoc network With maximum modularity The initial value of is 0, and each node is regarded as an independent cluster; According to the node In the channel The number of connections with other nodes within the communication range, computing the node In the channel The node degree on , expressed as ; in, is the total number of nodes, Representation Node With Node The connection relationship between nodes With Node Within communication range and with the same available channel, then ;otherwise, ; Get all nodes in the UAV ad hoc network on the channel The node degrees on the channel are averaged to obtain the node degrees of all nodes on the channel. The average node degree on , expressed as ; calculate and The absolute value of the difference is normalized and the node In the channel Adaptive node degree difference on , expressed as ; in, express and The absolute value of the difference, Represents a natural constant.

[0008] In one embodiment, calculating the node similarity with the introduction of multi-channel constraints and adaptive node degree differences includes: According to the multi-channel constraints of the UAV ad hoc network and the adaptive node degree difference of the node itself, the Jaccard similarity coefficient between any two nodes is corrected, and the node similarity with the introduction of multi-channel constraints and adaptive node degree difference is obtained. , expressed as ; in, is the total number of channels in the UAV ad hoc network, Representation Node In the channel Adaptive node degree difference on and Respectively represent nodes With Node In the channel The neighbor nodes on .

[0009] In one embodiment, the modularity of the drone ad hoc network is a modularity function based on node similarity, and its calculation formula is expressed as: ; in, is the sum of similarities of all nodes in the UAV ad hoc network, and Respectively represent nodes With Node The cluster to which it belongs, Representation Node With Node Belong to the same cluster If it belongs to, , otherwise ; and Respectively represent nodes With Node The sum of node similarities with all its neighbors; Representation Cluster The sum of similarities between nodes in For cluster The sum of similarities of the edges connecting the internal nodes; Indicates the proportion of intra-cluster connections, is the proportion of cluster connections to the outside world.

[0010] In one embodiment, the maximum modularity increment in the current cycle is added to the modularity of the UAV ad hoc network, and two clusters corresponding to the maximum modularity increment are merged, including: Increment the maximum modularity in the current loop Adding modularity to drone ad hoc networks In the above example, the modularity of the UAV ad hoc network is updated as , and merge the two clusters corresponding to the maximum modularity increment in the current cycle; If after updating, , the maximum modularity of the UAV self-organizing network Update to current module , save the current clustering result and enter the next cycle; otherwise, The value remains unchanged and enters the next cycle directly.

[0011] In one embodiment, the optimal clustering result of the drone self-organizing network includes: the number of clusters, the cluster size, the members in each cluster, and the cluster size deviation; wherein the number of clusters Indicates the total number of clusters in the drone network, the cluster size Indicates the number of drones in the network The number of members in a cluster, and ; Cluster size bias It is defined as the standard deviation of the cluster size, which is used to measure the balance of the number of members between clusters. It is specifically expressed as ; ; in, Indicates the average value of cluster size and the cluster size deviation The smaller the value of, the more balanced the distribution of cluster sizes.

[0012] A UAV self-organizing network clustering system based on adaptive node degree difference, the system comprising: The network topology structure building module is used to build the network topology structure of the UAV self-organizing network and obtain the connection edges between each node according to the available channel set and geographical location information perceived by each node in the UAV self-organizing network; The initialization setting and clustering module is used to initialize the modularity and clustering structure of the UAV self-organizing network, calculate the adaptive node degree difference of each node, and calculate the node similarity of introducing multi-channel constraints and adaptive node degree difference, perform clustering according to node similarity, and update the network topology structure of the UAV self-organizing network; The connection edge loop processing module is used to loop the connection edges between nodes in the UAV self-organizing network. In each loop, the number of connection edges decreases by 1 until the number of connection edges decreases to 0. In each loop, the modularity increment brought by merging the two clusters corresponding to each connection edge is traversed and calculated, and the connection edge with the largest modularity increment and a common channel after merging the clusters in the current loop is recorded. The modularity update and cluster merging module is used to add the maximum modularity increment in the current cycle to the modularity of the UAV ad hoc network and merge the two clusters corresponding to the maximum modularity increment; The clustering result output module is used to output the optimal clustering result of the UAV self-organizing network when the number of connection edges is reduced to 0.

[0013] Beneficial effects of the present application: The above-mentioned UAV self-organizing network clustering method and system based on adaptive node degree difference, by considering the multi-channel characteristics of large-scale UAV self-organizing networks, defines the node similarity of introducing multi-channel constraints and adaptive node degree difference for clustering. When clustering, the adaptive node degree difference can reflect the uniformity of node connection, making the clustering result more balanced, avoiding the occurrence of too many extremely large clusters or extremely small clusters, thereby causing excessive resource competition or waste problems; further, the connection edges between the nodes in the UAV self-organizing network after clustering are cyclically processed, and the modularity increment brought by the dynamic calculation of the merged clusters is selected to merge the clusters. The community structure and overall performance of the UAV network can be continuously optimized, the cluster scale deviation can be reduced, and the optimal clustering result can be achieved. This method is suitable for large-scale multi-channel UAV self-organizing network communication scenarios, and can dynamically organize large-scale UAV nodes to achieve efficient communication and resource management in a complex environment with limited spectrum resources and differentiated node available channels. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Figure 1 Schematic diagram of a process of clustering a UAV self-organizing network based on adaptive node degree difference in one embodiment; Figure 2 is a schematic diagram of a typical clustering network structure in an embodiment; Figure 3 A schematic diagram of a comparative experiment process in an embodiment; Figure 4 A schematic diagram of the initial stage of drone networking for 30 nodes in one embodiment; Figure 5A schematic diagram of cluster network construction for a 30-node drone network in one embodiment; Figure 6 A schematic diagram of the initial stage of drone networking for 50 nodes in one embodiment; Figure 7 A schematic diagram of building a clustered network for a 50-node drone network in one embodiment; Figure 8 A schematic diagram of the initial stage of drone networking with 100 nodes in one embodiment; Fig. 9 A schematic diagram of building a clustered network for a 100-node drone network in one embodiment; Fig.10 A schematic diagram of the effect of the number of nodes on the average cluster size deviation according to an embodiment; Fig.11 A schematic diagram of the effect of the number of channels on the average cluster size deviation in one embodiment; Fig.12 A schematic diagram of the effect of the number of nodes on network throughput in one embodiment; Fig.13 FIG. 1 is a schematic diagram showing the effect of the number of channels on network throughput in an embodiment. DETAILED DESCRIPTION

[0015] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0016] In one embodiment, Figure 1 As shown, a method for clustering a UAV self-organizing network based on adaptive node degree difference is provided, comprising the following steps: Step S1, based on the available channel set and geographic location information perceived by each node in the drone self-organizing network, the network topology structure of the drone self-organizing network is constructed and the connection edges between the nodes are obtained.

[0017] Among them, the network topology structure of the drone self-organizing network constructed in this application is a typical clustered network structure, such as Figure 2 Shown, including nodes, the spectrum is divided into Non-overlapping channels, each node is equipped with a half-duplex transceiver for sensing channel availability. Any two nodes can communicate directly if they are within the communication range of each other and have the same available channel. The communication distance between nodes is limited by the maximum communication distance. .

[0018] Step S2, initialize the modularity and clustering structure of the UAV self-organizing network, calculate the adaptive node degree difference of each node, and calculate the node similarity of introducing multi-channel constraints and adaptive node degree difference, cluster according to node similarity, and update the network topology structure of the UAV self-organizing network.

[0019] Specifically, set the modularity of the UAV self-organizing network With maximum modularity The initial value of is 0, and each node is regarded as an independent cluster; According to the node In the channel The number of connections with other nodes within the communication range, computing the node In the channel The node degree on , expressed as ; in, is the total number of nodes, Representation Node With Node The connection relationship between nodes With Node Within communication range and with the same available channel, then ;otherwise, ; Get all nodes in the UAV ad hoc network on the channel The node degrees on the channel are averaged to obtain the node degrees of all nodes on the channel. The average node degree on , expressed as ; calculate and The absolute value of the difference is normalized and the node In the channel Adaptive node degree difference on , expressed as ; in, express and The absolute value of the difference, Represents a natural constant.

[0020] The adaptive node degree difference can measure the uniformity of node connections and the difference of node connections relative to the overall distribution of the network. The smaller the adaptive node degree difference, the closer the connection status of the node is to the average level of the network, reflecting the uniformity and importance of the node in the network.

[0021] Furthermore, the Jaccard similarity coefficient between any two nodes is modified according to the multi-channel constraint conditions of the UAV ad hoc network and the adaptive node degree difference of the node itself, and the node similarity with the introduction of multi-channel constraint conditions and adaptive node degree difference is obtained. , expressed as ; in, is the total number of channels in the UAV ad hoc network, Representation Node In the channel Adaptive node degree difference on and Respectively represent nodes With Node In the channel The neighbor nodes on .

[0022] Step S3, loop processing the connection edges between each node in the UAV self-organizing network, and the number of connection edges is reduced by 1 in each loop until the number of connection edges is reduced to 0; wherein, in each loop, the modularity increment brought by merging the two clusters corresponding to each connection edge is traversed and calculated, and the connection edge with the largest modularity increment and the existence of a common channel after merging the clusters in the current loop is recorded.

[0023] Specifically, the modularity of the UAV self-organizing network is a modularity function based on node similarity, and its calculation formula is expressed as: ; in, is the sum of similarities of all nodes in the UAV ad hoc network, and Respectively represent nodes With Node The cluster to which it belongs, Representation Node With Node Are they in the same cluster? If it belongs to, , otherwise ; and Respectively represent nodes With Node The sum of node similarities with all its neighbors; Representation Cluster The sum of similarities between nodes in For cluster The sum of similarities of the edges connecting the internal nodes; Indicates the proportion of intra-cluster connections, is the proportion of cluster connections to the outside world.

[0024] The modularity function is currently widely used as a standard for community quality evaluation. The calculation formula of the traditional modularity function is: ; in, is the sum of all values, Is a node With Node The weight of the edge between them; Representation and Node The sum of the weights of the connected edges, Representation and Node The sum of the weights of the connected edges, and Respectively represent nodes With Node The community or cluster to which the user belongs.

[0025] The traditional modularity function mainly evaluates the clustering quality of the network by maximizing the proportion of intra-cluster connections and minimizing the proportion of inter-cluster connections. The higher the modularity, the tighter the intra-cluster connection. However, if the cluster size balance is not considered, it is easy to cause too many extremely large clusters or extremely small clusters, thereby causing excessive resource competition or waste. To this end, this application introduces node similarity that combines multi-channel constraints and adaptive node degree differences, and constructs a modularity function based on node similarity, so that clustering not only considers the tight intra-cluster connection, but also achieves a more balanced cluster size, avoiding the situation where extremely large clusters are crowded and lead to intensified resource competition and serious collisions.

[0026] Step S4, adding the maximum modularity increment in the current cycle to the modularity of the UAV ad hoc network, and merging the two clusters corresponding to the maximum modularity increment.

[0027] Specifically, the maximum modularity increment in the current cycle is Adding modularity to drone ad hoc networks In the above example, the modularity of the UAV ad hoc network is updated as , and merge the two clusters corresponding to the maximum modularity increment in the current cycle; If after updating, , the maximum modularity of the UAV self-organizing network Update to current module , save the current clustering result and enter the next cycle; otherwise, The value remains unchanged and enters the next cycle directly.

[0028] Step S5: when the number of connected edges is reduced to 0, the optimal clustering result of the UAV self-organizing network is output.

[0029] Specifically, the optimal clustering results of the UAV self-organizing network include: the number of clusters, the cluster size, the members in each cluster, and the cluster size deviation; among them, the number of clusters Indicates the total number of clusters in the drone network, the cluster size Indicates the number of drones in the network The number of members in a cluster, and ; Cluster size bias It is defined as the standard deviation of the cluster size, which is used to measure the balance of the number of members between clusters. It is specifically expressed as ; ; in, Indicates the average value of cluster size and the cluster size deviation The smaller the value of, the more balanced the distribution of cluster size. In UAV ad hoc networks, cluster size balance is an important condition for network clustering. A relatively balanced number of members in each cluster not only facilitates cluster head management, but also improves the reuse efficiency of communication resources.

[0030] The above-mentioned UAV self-organizing network clustering method based on adaptive node degree difference introduces multi-channel constraints and node similarity with adaptive node degree difference for clustering, and uses modularity based on node similarity for cluster optimization merging to achieve efficient and balanced clustering results. It can dynamically organize large-scale UAV nodes to achieve efficient communication and resource management in a complex environment with limited spectrum resources and differentiated node available channels.

[0031] In order to verify the beneficial effect of the above-mentioned UAV self-organizing network clustering method based on adaptive node degree difference, a comparative experimental process was further carried out. The comparative experimental process is basically consistent with the implementation process of the method proposed in this application. Only clustering is performed according to node similarity, and four node similarities are traversed. ,like Figure 3 As shown. Similarity of different nodes Different clustering strategies will form different topological structures, including: : Unweighted similarity, by judging whether two nodes are neighbors of each other (within the communication range and with a common channel); : Jaccard neighbor similarity, the ratio of the number of common neighbor nodes between two nodes to the total number of neighbor nodes; : Add multi-channel constraints based on Jaccard similarity.

[0032] : The Jaccard similarity with multi-channel constraints and adaptive node degree difference is the node similarity proposed in this application.

[0033] During the experiment, the network simulation scenario was set to randomly deploy drone nodes in The maximum communication radius of the node is 200m, and the number of network nodes is , the number of available channels ranges from ,The available channel of each node is randomly selected from the total channels.,For different clustering strategies, the simulation uses the same node location information and ,available channel information to conduct experiments.,Multiple experiments are conducted under each clustering strategy, and the average value is taken to ensure the ,reliability of the results.

[0034] For intuitive display, 30, 50, and 100 nodes are randomly deployed in the rectangular area, and the total number of available channels is 6. The initial stage of UAV networking is as follows: Figure 4 , Figure 6 , Figure 8 shown. Figure 5 , Figure 7 , Fig. 9 The schematic diagram of cluster network construction for UAV networking is shown.

[0035] First, we evaluate the effect of the number of nodes on the average cluster size deviation when the total number of channels is fixed: Set the total number of available channels in the network to 6 and the number of nodes to The effect of the number of nodes on the average cluster size deviation is as follows: Fig.10 As shown, it can be seen that with the increase in the number of nodes, the network complexity rises, the randomness of node distribution and the channel resource competition become more significant, resulting in a more unbalanced distribution of cluster sizes and a significant increase in cluster size deviation.

[0036] different The values ​​correspond to different cluster formation strategies. Specifically, Assuming that nodes are within the communication range and have a common channel, they can form a cluster. This has high flexibility, but the cluster size deviation is the largest, making it difficult to effectively control the formation of clusters. On this basis, adding constraints on the similarity of neighbor nodes reduces the cluster size deviation, but does not consider more complex network environments. The requirement for consistency of channel conditions is increased. Nodes must have a common channel to form a cluster, which makes cluster formation more restricted and reduces deviation. We further introduce adaptive node degree difference, dynamically adjust the cluster formation according to the adaptive node degree difference, which is more adaptable and can maintain a more balanced topological structure under different node number distribution. Compared with the other three strategies, The cluster size deviation is significantly reduced, indicating that this application has better performance on the balance of network topology structure.

[0037] Secondly, we evaluate the impact of the total number of channels on the average cluster size deviation when the number of nodes is fixed: Assume that the number of nodes in the network is 150 and the number of channels is The influence of the number of channels on the average cluster size deviation is shown in Fig.11 As shown in Figure 2, it can be seen that as the number of channels increases, the cluster size deviation gradually increases. This is because more channel selections reduce the probability of sharing channels between nodes, resulting in more dispersed cluster formation, which in turn increases the cluster size difference and the average cluster size deviation. Under different channel numbers, The cluster size deviation of the clustering strategy is kept to the minimum, and the size distribution of the cluster structure is more balanced.

[0038] Finally, the network throughput was evaluated. After the randomly distributed drone nodes were clustered into a network structure according to the method proposed in this application, the network saturation throughput was statistically analyzed using the Bianchi model to analyze the network performance. This model accurately describes the throughput performance of nodes using the CSMA / CA (Carrier Sense Multiple Access with Collision Avoid) protocol under limited node numbers and ideal channel conditions by establishing a two-dimensional Markov chain.

[0039] (1) When the total number of channels is fixed, the impact of the change in the number of nodes on the network throughput: In order to verify the impact of different clustering strategies on network throughput when the number of nodes changes, the total number of channels is set to 6 and the number of nodes is The impact of the number of nodes on network throughput is as follows: Fig.12 shown.

[0040] Depend on Fig.12 It can be seen that the clustering strategy proposed in this application The cluster size deviation is minimal under different node numbers. This application reduces the difference in the number of members in the cluster by balancing the cluster size, avoiding the situation where the cluster is too large or too small, making the channel competition under the CSMA / CA mechanism fairer and effectively improving resource utilization and network throughput. The network throughput of the clustering strategy is significantly higher than that of the other three strategies.

[0041] (2) When the total number of nodes is fixed, the impact of changes in the number of channels on network throughput: In order to verify the impact of different clustering strategies on throughput when the total number of channels changes, the total number of nodes is set to 150 and the total number of channels is set to The influence of the number of channels on the network throughput is as follows: Fig.13 As shown. Fig.13 Overall, under different total channel numbers, the clustering strategy proposed in this application The network throughput of the is higher than that of the other three, which shows that the network with smaller cluster size deviation can make full use of channel resources and improve the network throughput.

[0042] To sum up, the comparative test results show that the drone self-organizing network clustering method based on adaptive node degree difference proposed in this application can form a more balanced network topology structure, significantly reduce the cluster scale deviation, effectively improve the network throughput, have good scalability, and can support large-scale network communications in complex environments. It can be used as an effective supplement to the current mainstream multi-channel drone networking method.

[0043] In one embodiment, a UAV self-organizing network clustering system based on adaptive node degree difference is provided, comprising: The network topology structure building module is used to build the network topology structure of the UAV self-organizing network and obtain the connection edges between each node according to the available channel set and geographical location information perceived by each node in the UAV self-organizing network; The initialization setting and clustering module is used to initialize the modularity and clustering structure of the UAV self-organizing network, calculate the adaptive node degree difference of each node, and calculate the node similarity of introducing multi-channel constraints and adaptive node degree difference, perform clustering according to node similarity, and update the network topology structure of the UAV self-organizing network; The connection edge loop processing module is used to loop the connection edges between nodes in the UAV self-organizing network. In each loop, the number of connection edges decreases by 1 until the number of connection edges decreases to 0. In each loop, the modularity increment brought by merging the two clusters corresponding to each connection edge is traversed and calculated, and the connection edge with the largest modularity increment and a common channel after merging the clusters in the current loop is recorded. The modularity update and cluster merging module is used to add the maximum modularity increment in the current cycle to the modularity of the UAV ad hoc network and merge the two clusters corresponding to the maximum modularity increment; The clustering result output module is used to output the optimal clustering result of the UAV self-organizing network when the number of connection edges is reduced to 0.

[0044] For the specific definition of the unmanned aerial vehicle self-organizing network clustering system based on adaptive node degree difference, please refer to the definition of the unmanned aerial vehicle self-organizing network clustering method based on adaptive node degree difference in the above text, which will not be repeated here. Each module in the above-mentioned unmanned aerial vehicle self-organizing network clustering system based on adaptive node degree difference can be implemented in whole or in part by software, hardware and a combination thereof. The above-mentioned modules can be embedded in or independent of the processor in the computer device in the form of hardware, or can be stored in the memory of the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.

[0045] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0046] The above-described embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the present application. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the attached claims.

Claims

1. A clustering method for unmanned aerial vehicle ad hoc networks based on adaptive node degree difference, characterized in that: The method comprises: According to the available channel set and geographic location information perceived by each node in the UAV self-organizing network, the network topology of the UAV self-organizing network is constructed and the connection edges between each node are obtained; Initialize the modularity and clustering structure of the UAV self-organizing network, calculate the adaptive node degree difference of each node, and calculate the node similarity of introducing multi-channel constraints and adaptive node degree difference, perform clustering according to the node similarity, and update the network topology structure of the UAV self-organizing network; The connection edges between nodes in the UAV self-organizing network are processed cyclically. The number of connection edges is reduced by 1 in each cycle until the number of connection edges is reduced to 0. In each cycle, the modularity increment brought by merging the two clusters corresponding to each connection edge is calculated and recorded. The connection edge with the largest modularity increment and a common channel after merging the clusters in the current cycle is recorded. Add the maximum modularity increment in the current cycle to the modularity of the UAV ad hoc network, and merge the two clusters corresponding to the maximum modularity increment; When the number of connected edges is reduced to 0, the optimal clustering result of the UAV self-organizing network is output.

2. The method according to claim 1, characterized in that The network topology structure of the UAV self-organizing network includes nodes, the spectrum is divided into Non-overlapping channels, each node is equipped with a half-duplex transceiver for sensing channel availability. Any two nodes can communicate directly if they are within the communication range of each other and have the same available channel. The communication distance between nodes is limited by the maximum communication distance. .

3. The method according to claim 1, characterized in that Initialize the modularity and clustering structure of the UAV self-organizing network, and calculate the adaptive node degree difference of each node, including: Set the modularity of the drone ad hoc network With maximum modularity The initial value of is 0, and each node is regarded as an independent cluster; According to the node In the channel The number of connections with other nodes within the communication range, computing the node In the channel The node degree on , expressed as ； in, is the total number of nodes, Representation Node With Node The connection relationship between nodes With Node Within communication range and with the same available channel, then ;otherwise, ; Get all nodes in the UAV ad hoc network on the channel The node degrees on the channel are averaged to obtain the node degrees of all nodes on the channel. The average node degree on , expressed as ； calculate and The absolute value of the difference is normalized and the node In the channel Adaptive node degree difference on , expressed as ； in, express and The absolute value of the difference, Represents a natural constant.

4. The method according to claim 3, characterized in that Calculate the node similarity with the introduction of multi-channel constraints and adaptive node degree difference, including: According to the multi-channel constraints of the UAV ad hoc network and the adaptive node degree difference of the node itself, the Jaccard similarity coefficient between any two nodes is corrected, and the node similarity with the introduction of multi-channel constraints and adaptive node degree difference is obtained. , expressed as ； in, is the total number of channels in the UAV ad hoc network, Representation Node In the channel Adaptive node degree difference on and Respectively represent nodes With Node In the channel The neighbor nodes on .

5. The method according to claim 1 or 3, characterized in that: Modularity of UAV Ad Hoc Networks is a modularity function based on node similarity, and its calculation formula is expressed as: ； in, is the sum of similarities of all nodes in the UAV ad hoc network, and Respectively represent nodes With Node The cluster to which it belongs, Representation Node With Node Are they in the same cluster? If it belongs to, , otherwise ; and Respectively represent nodes With Node The sum of node similarities with all its neighbors; Representation Cluster The sum of similarities between nodes in For cluster The sum of similarities of the edges connecting the internal nodes; Indicates the proportion of intra-cluster connections, is the proportion of cluster connections to the outside world.

6. The method according to claim 5, characterized in that The maximum modularity increment in the current cycle is added to the modularity of the UAV ad hoc network, and the two clusters corresponding to the maximum modularity increment are merged, including: Increment the maximum modularity in the current loop Adding modularity to drone ad hoc networks In the above example, the modularity of the UAV ad hoc network is updated as , and merge the two clusters corresponding to the maximum modularity increment in the current cycle; If after updating, , the maximum modularity of the UAV self-organizing network Update to current module , save the current clustering result and enter the next cycle; otherwise, The value remains unchanged and enters the next cycle directly.

7. The method according to claim 1, characterized in that The optimal clustering results of UAV self-organizing network include: number of clusters, cluster size, members in each cluster and cluster size deviation; among them, the number of clusters Indicates the total number of clusters in the drone network, the cluster size Indicates the number of drones in the network The number of members in a cluster, and ; Cluster size bias It is defined as the standard deviation of the cluster size, which is used to measure the balance of the number of members between clusters. It is specifically expressed as ； ； in, Indicates the average value of cluster size and the cluster size deviation The smaller the value of, the more balanced the distribution of cluster sizes.

8. A UAV self-organizing network clustering system based on adaptive node degree difference, characterized in that: The system comprises: The network topology structure building module is used to build the network topology structure of the UAV self-organizing network and obtain the connection edges between each node according to the available channel set and geographical location information perceived by each node in the UAV self-organizing network; The initialization setting and clustering module is used to initialize the modularity and clustering structure of the UAV self-organizing network, calculate the adaptive node degree difference of each node, and calculate the node similarity of the introduction of multi-channel constraints and the adaptive node degree difference, perform clustering according to the node similarity, and update the network topology structure of the UAV self-organizing network; The connection edge loop processing module is used to loop the connection edges between nodes in the UAV self-organizing network. In each loop, the number of connection edges decreases by 1 until the number of connection edges decreases to 0. In each loop, the modularity increment brought by merging the two clusters corresponding to each connection edge is traversed and calculated, and the connection edge with the largest modularity increment and a common channel after merging the clusters in the current loop is recorded. The modularity update and cluster merging module is used to add the maximum modularity increment in the current cycle to the modularity of the UAV ad hoc network and merge the two clusters corresponding to the maximum modularity increment; The clustering result output module is used to output the optimal clustering result of the UAV self-organizing network when the number of connection edges is reduced to 0.

Images