A multi-hop clustering method to improve the topology duration of aeronautical ad hoc networks

Abstract

A multi-hop clustering method for improving the topology duration of an aviation ad hoc network. It includes a clustering formation stage and a clustering maintenance stage. The clustering formation stage includes: preprocessing flight trajectory data; broadcasting data; generating a neighbor attribute table; calculating link duration; counting the number of links greater than a link duration threshold; when the number of links greater than the link duration threshold exceeds a set minimum connectivity value, the node submits a link establishment request to a neighbor node to form a network cluster. The clustering maintenance stage includes: updating trajectory information in the neighbor attribute table; calculating the link duration of existing links and the link duration between new neighbor nodes based on the trajectory attributes of neighbor nodes in the neighbor attribute table; maintaining existing connections and creating new connections based on the calculated link duration to maintain clustering, etc. Advantages of the present invention: clustering is established in a distributed manner; link duration can be effectively improved.

Description

Technical Field

[0001] The invention belongs to the technical field of aviation communications, and in particular relates to a multi-hop clustering method for improving the topology duration of an aviation ad hoc network. Background Art

[0002] The aviation communication system is the core infrastructure to ensure the efficient operation of the aviation transportation system, so it is necessary to ensure the stability and timeliness of data transmission. Aeronautical ad hoc network is one of the effective solutions for aviation communication system, which relies on aircraft nodes as relays to transmit data through multi-hop air-to-air communication links. However, the high-speed movement of nodes in the aeronautical ad hoc network will lead to highly dynamic changes in the network topology of the aeronautical ad hoc network. The highly dynamic network topology will cause the routing information of network nodes to frequently fail, resulting in reduced data transmission efficiency or even transmission failure.

[0003] At present, the aviation ad hoc network routing protocol is still improved on the traditional mobile ad hoc network routing protocol. However, these methods have many problems: 1) These routing protocols need to consume a lot of network resources to frequently update routing information to adapt to the highly dynamic network topology; 2) These flat routing protocols will cause the resource cost of maintaining routing information to increase exponentially as the network scale continues to increase. Therefore, it is necessary to study a sustainable method for network topology.

[0004] Aerial MANET clustering refers to placing relatively stable nodes in the same cluster to form a subnet based on their location and mobility attributes. Currently, a few scholars have used methods such as K-means, DBSCAN and learning vector quantization in Aerial MANET clustering to improve the sustainability of Aerial MANET network topology. However, since these methods only perform a simple weighted combination of the location attributes and mobility attributes of network nodes, these methods have limited effect on improving the sustainability of network topology. Summary of the invention

[0005] In order to solve the problem of low sustainability of network topology caused by high dynamic changes in the network topology of an aeronautical ad hoc network, the purpose of the present invention is to provide a method that can fully utilize the location attributes and mobility attributes of the nodes in the aeronautical ad hoc network to form a stable network clustering structure, that is, a multi-hop clustering method that improves the duration of the aeronautical ad hoc network topology. Its goal is to use the link duration as a clustering indicator to establish a stable air-to-air link between nodes to improve the duration of the aeronautical ad hoc network topology.

[0006] In order to achieve the above object, the multi-hop clustering method for improving the topology duration of an aeronautical ad hoc network provided by the present invention comprises the following steps performed in sequence:

[0007] (1) Cluster formation stage

[0008] 1.1) Preprocess the original data of the flight trajectory of the aviation node to obtain the preprocessed trajectory attributes;

[0009] 1.2) Each node sends the pre-processed trajectory attributes to neighboring nodes in a broadcast manner;

[0010] 1.3) Each node obtains the trajectory attributes of all its neighbor nodes and generates a neighbor attribute table;

[0011] 1.4) Calculate the link duration between the node and its neighbor nodes according to the trajectory attributes in the neighbor attribute table;

[0012] 1.5) Count the number of links m that are greater than the link duration threshold;

[0013] 1.6) When the number of links m that are greater than the link duration threshold exceeds the set minimum connectivity n, the node makes a link establishment request to the neighboring node to form a network cluster;

[0014] (2) Cluster maintenance phase

[0015] 2.1) Obtain the trajectory attributes of the neighbor nodes and update the trajectory information of the neighbor nodes in the neighbor attribute table;

[0016] 2.2) Calculating the link duration of the existing link and the link duration between the new neighbor node according to the trajectory attributes of the neighbor node in the neighbor attribute table;

[0017] 2.3) maintaining the existing connections and creating new connections according to the link duration calculated in step 2.2) to maintain clustering;

[0018] 2.4) Repeat steps 2.1) to 2.3) to cluster the dynamic network in real time according to the trajectory information of the nodes.

[0019] In step 1.1), the method of preprocessing the original data of the flight trajectory of the aviation node to obtain the preprocessed trajectory attributes is:

[0020] The original data of the flight trajectory of the aviation node is converted from the WGS-84 coordinate system (World Geodetic System-1984 Coordinate System) to the Earth-Centered Earth-Fixed (ECEF) coordinate system. The conversion formula is shown in formula (1):

[0021]

[0022] Where B, L and H represent the latitude, longitude and altitude in the WGS-84 coordinate system respectively; x, y, z represent the ECEF coordinate position respectively; a represents the semi-major axis of the earth, which is 6,378,137m; b represents the semi-minor axis of the earth, which is 6,356,752m; N R represents the main vertical radius of curvature; e represents the first eccentricity of the ellipsoid.

[0023] In step 1.2), the method in which each node sends the pre-processed trajectory attributes to neighboring nodes in a broadcasting manner is:

[0024] The preprocessed trajectory attributes are placed in a data packet in a fixed format, and the current node is used as the source node, and the data packet is sent to all one-hop neighbor nodes by broadcasting.

[0025] In step 1.3), each node obtains the trajectory attributes of all its neighbor nodes, and the method for generating the neighbor attribute table is:

[0026] When a node broadcasts its own attributes, it will receive trajectory attribute data packets from neighboring nodes. The node collects the trajectory attributes of all neighboring nodes to generate a neighbor attribute table.

[0027] In step 1.4), the method for calculating the link duration between a node and its neighbor based on the trajectory attributes of the neighbor node is:

[0028] According to the node location attribute and node mobility attribute, the link duration between nodes is calculated; link duration refers to the duration that two nodes can communicate with each other, and the calculation formula is shown in formula (2):

[0029]

[0030] Among them, Stime ij represents the duration of time that nodes i and j are within its transmission range; D ij Indicates the relative position of node i and node j; V ij represents the relative speed between node i and node j; R represents the communication radius between node i and node j.

[0031] In step 1.5), the method for counting the number m of links greater than the link duration threshold is:

[0032] The link duration threshold is a hyperparameter in this method and needs to be set in advance. It represents the minimum value of the duration required to establish a link between nodes. The number of links greater than the link duration threshold is the number of neighbors in the neighbor attribute table that meet the conditions for establishing a link between nodes.

[0033] In step 1.6), the method of forming a network cluster where, when the number of links greater than the link duration threshold exceeds the set minimum connectivity n, the node sends a link establishment request to its neighbor nodes is as follows:

[0034] The minimum connectivity n here is another hyperparameter in this method and also needs to be set in advance. From step 1.5), the number of neighbors that meet the condition for establishing a link between nodes can be obtained. When the number of neighbors m of this node that meet the condition for establishing a link between nodes is greater than the minimum connectivity n, a connection establishment request is sent to its neighbor nodes to establish a link between nodes; otherwise, the node waits for connection requests from other nodes.

[0035] In step 2.1), the method of updating the trajectory information of neighbor nodes in the neighbor attribute table is as follows:

[0036] This process is the same as steps 1.1) to 1.3) in the cluster formation stage. First, each node preprocesses its original flight trajectory data to obtain the preprocessed trajectory attributes, then broadcasts the preprocessed trajectory attributes and receives the trajectory attributes of its neighbors at the same time, and then updates the neighbor attribute table according to the received trajectory attributes.

[0037] In step 2.2), the method of calculating the link duration of existing links and the link duration between the node and a new neighbor node according to the trajectory attributes of neighbor nodes in the neighbor attribute table is as follows:

[0038] This process is the same as step 1.4) in the cluster formation stage. According to the position attributes and movement attributes of the nodes, the link duration between nodes is calculated using formula (2).

[0039] In step 2.3), the method of maintaining existing connections and creating new connections to maintain the cluster according to the link duration calculated in step 2.2) is as follows:

[0040] First, count the number of existing connections n1 greater than the link duration threshold, the number of existing connections n2 less than the link duration threshold, and the number of new connections n3 greater than the link duration threshold respectively; when n1>0 and n1 + n3≥n, the existing connections remain unchanged and new connections are established; when n1>0 and n1 + n3<n, only the existing connections remain unchanged; when n 1＝ 0 and n2 + n3≥n, the existing connections remain unchanged and new connections are established; when n1 = 0 and n2 + n3<n, all connections are disconnected and the node waits for a link establishment request from other nodes.

[0041] In step 2.4), the method of repeating steps 2.1) to step 2.3) to cluster the dynamic network in real time according to the node trajectory information is:

[0042] After each time period, step 2.1), step 2.2) and step 2.3) are sequentially executed to form a real-time and effective network clustering, thereby improving the continuity of the network topology.

[0043] Compared with the existing clustering algorithms, the multi-hop clustering method for improving the topology duration of an aviation ad hoc network provided by the present invention has the following advantages: 1) clustering is established in a distributed manner; 2) link duration can be effectively improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 A flow chart of the clustering formation phase in the multi-hop clustering method for improving the topology duration of an aviation ad hoc network provided by the present invention.

[0045] Figure 2 A flow chart of the clustering maintenance phase in the multi-hop clustering method for improving the topology duration of an aviation ad hoc network provided by the present invention.

[0046] Figure 3 It is the node distribution diagram of a certain time slice.

[0047] Figure 4 It is the moving trajectory diagram of all nodes. DETAILED DESCRIPTION

[0048] The multi-hop clustering method for improving the topology duration of an aeronautical ad hoc network provided by the present invention is described in detail below with reference to the accompanying drawings and specific implementation examples.

[0049] like Figure 1 and Figure 2 As shown, the multi-hop clustering method for improving the topology duration of an aeronautical ad hoc network provided by the present invention comprises the following steps performed in sequence:

[0050] (1) Cluster formation stage

[0051] 1.1) Preprocess the original data of the flight trajectory of the aviation node to obtain the preprocessed trajectory attributes;

[0052] The original data of the flight trajectory of the aviation node is converted from the WGS-84 coordinate system (World Geodetic System-1984 Coordinate System) to the Earth-Centered Earth-Fixed (ECEF) coordinate system. The conversion formula is shown in formula (1):

[0053]

[0054] Where B, L and H represent the latitude, longitude and altitude in the WGS-84 coordinate system. x, y and z represent the ECEF coordinate position. A represents the semi-major axis of the earth, which is 6,378,137m. b represents the semi-minor axis of the earth, which is 6,356,752m. R represents the principal vertical radius of curvature. e represents the first eccentricity of the ellipsoid.

[0055] 1.2) Each node sends the pre-processed trajectory attributes to neighboring nodes in a broadcast manner;

[0056] The preprocessed trajectory attributes are placed in a data packet in a fixed format, and the current node is used as the source node, and the data packet is sent to all one-hop neighbor nodes by broadcasting.

[0057] 1.3) Each node obtains the trajectory attributes of all its neighbor nodes and generates a neighbor attribute table;

[0058] When a node broadcasts its own attributes, it will receive trajectory attribute data packets from neighboring nodes. The node collects the trajectory attributes of all neighboring nodes to generate a neighbor attribute table.

[0059] 1.4) Calculate the link duration between the node and its neighbor nodes according to the trajectory attributes of the neighbor nodes in the neighbor attribute table;

[0060] According to the location attribute and mobility attribute of the node, the link duration between nodes is calculated. Link duration refers to the duration that two nodes can communicate with each other. The calculation formula is shown in formula (2):

[0061]

[0062] Among them, Stime ij represents the duration of time that nodes i and j are within their transmission range; D ij represents the relative position of node i and node j, V ij represents the relative speed between node i and node j, and R represents the communication radius between node i and node j.

[0063] 1.5) Count the number of links m that are greater than the link duration threshold;

[0064] The link duration threshold is a hyperparameter of the method (according to experiments, the present invention sets the link duration threshold to 600s), which needs to be set in advance and represents the minimum value of the duration of the link between nodes. Counting the number of links m greater than the link duration threshold refers to counting the number of neighbors in the neighbor attribute table that meet the conditions for establishing a link between nodes.

[0065] 1.6) When the number m of links greater than the link duration threshold exceeds the set minimum connectivity n, the node sends a link establishment request to its neighbor nodes to form a network cluster.

[0066] The minimum connectivity n here is another hyperparameter in this method (according to experiments, the present invention sets the minimum connectivity n to 2), which also needs to be set in advance. From step 1.5), the number of neighbors that meet the condition for establishing a link between nodes can be obtained. When the node meets the condition that the number of neighbors is greater than the minimum connectivity n, it sends a connection establishment request to its neighbor nodes to establish a link between nodes; otherwise, the node waits for connection requests from other nodes.

[0067] (2) Cluster maintenance phase

[0068] 2.1) Obtain the trajectory attributes of neighbor nodes and update the trajectory information of neighbor nodes in the neighbor attribute table;

[0069] This process is the same as steps 1.1) to 1.3) in the cluster formation phase. First, each node preprocesses its original flight trajectory data, then broadcasts the preprocessed trajectory attributes and receives the trajectory attributes of its neighbors at the same time, and then updates the neighbor attribute table according to the received trajectory attributes.

[0070] 2.2) Calculate the link duration of existing links and the link duration between the node and new neighbor nodes according to the trajectory attributes of neighbor nodes in the neighbor attribute table;

[0071] This process is the same as step 1.4) in the cluster formation phase. According to the position attribute and movement attribute of the node, the link duration between nodes is calculated using formula (2).

[0072] 2.3) Maintain existing connections and create new connections according to the link duration calculated in step 2.2) to maintain the cluster;

[0073] First, count the number n1 of existing connections greater than the link duration threshold, the number n2 of existing connections less than the link duration threshold, and the number n3 of new connections greater than the link duration threshold respectively. When n1>0 and n1 + n3≥n, keep the existing connections unchanged and establish new connections; when n1>0 and n1 + n3 < n, only keep the existing connections unchanged; when n1 = 0 and n2 + n3≥n, keep the existing connections unchanged and establish new connections; when n1 = 0 and n2 + n3 < n, disconnect all connections and wait for requests from other nodes to establish links.

[0074] 2.4) Repeat steps 2.1) to 2.3) to perform network clustering on the dynamic network in real time according to the node trajectory information.

[0075] After each time period, step 2.1), step 2.2) and step 2.3) are sequentially executed to form a real-time and effective network clustering, thereby improving the continuity of the network topology.

[0076] The effect of the multi-hop clustering method for improving the topology duration of an aeronautical ad hoc network provided by the present invention can be further illustrated by the following experimental results.

[0077] Evaluation index description: In order to quantitatively evaluate the method of the present invention, the following three evaluation indicators are used: average sustained time of link (AvgStime link ), Number of link changes (Number of link changes, Num Clink ), Average number of orphaned nodes (Average number of orphaned nodes, AvgNumO node ).

[0078]

[0079]

[0080]

[0081] Where n represents the total number of connections established between nodes, Stime i represents the duration of the ith connection, timeslice represents the number of time slices in the experimental process, and Clink j Indicates the number of disconnections in time slice j, NumO j Indicates the number of isolated nodes that have no connection established in time slice j.

[0082] In order to illustrate the effectiveness of the method of the present invention, the method of the present invention is compared with the DBSCAN algorithm. The DBSCAN algorithm simply weights the trajectory attributes of the nodes to perform network clustering. The method of the present invention is inspired by the DBSCAN algorithm and introduces the concept of link duration to establish clustering. Therefore, the comparison with the DBSCAN algorithm can verify the effectiveness of the present invention.

[0083] DBSCAN algorithm description:

[0084] DBSCAN algorithm. A classic clustering algorithm. References: M.Shahbazi, M.imsek and B.Kantarci, "Density-Based Clustering and Performance Enhancement oferonautical Ad Hoc Networks," 2022 International Balkan Conference on Communications and Networking (BalkanCom), 2022, pp.51-56, doi:10.1109 / BalkanCom55633.2022.9900681.

[0085] In order to illustrate the effectiveness of the method of the present invention for clustering nodes in aviation ad hoc networks, historical aviation flight data over the United States from 14:00 to 15:00 on June 27, 2022 was obtained from the OpenSky Network, and 571 nodes were extracted from the original data. The node distribution diagram of a certain time slice is shown in the figure below: Figure 3 The movement trajectory of all nodes in the entire time period is shown as Figure 4 shown.

[0086] It can be seen from Table 1 that the method of the present invention is significantly better than the DBSCAN algorithm in terms of the average number of isolated nodes, the number of link changes and the average link duration.

[0087] Table 1 Comparison of experimental results

[0088]

Claims

1. A multi-hop clustering method for improving the duration of topology of an aeronautical ad hoc network, characterized in that: the multi-hop clustering method for improving the duration of topology of an aeronautical ad hoc network comprises the following steps performed in sequence: (1) Cluster formation stage 1.1) Preprocess the original data of the flight trajectory of the aviation node to obtain the preprocessed trajectory attributes; 1.2) Each node sends the pre-processed trajectory attributes to neighboring nodes in a broadcast manner; 1.3) Each node obtains the trajectory attributes of all its neighbor nodes and generates a neighbor attribute table; 1.4) Calculate the link duration between the node and its neighbor nodes according to the trajectory attributes in the neighbor attribute table; 1.5) Count the number of links m that are greater than the link duration threshold; 1.6) When the number of links m that are greater than the link duration threshold exceeds the set minimum connectivity n, the node makes a link establishment request to the neighboring node to form a network cluster; (2) Cluster maintenance phase 2.1) Obtain the trajectory attributes of the neighbor nodes and update the trajectory information of the neighbor nodes in the neighbor attribute table; 2.2) Calculating the link duration of the existing link and the link duration between the new neighbor node according to the trajectory attributes of the neighbor node in the neighbor attribute table; 2.3) Maintain existing connections and create new connections according to the link duration calculated in step 2.2) to maintain clustering; first, count the number of existing connections greater than the link duration threshold as n1, the number of existing connections less than the link duration threshold as n2, and the number of new connections greater than the link duration threshold as n3; when n1 > 0 and n1 + n3 ≥ n, keep the existing connections unchanged and establish new connections; when n1 > 0 and n1 + n3 < n, only keep the existing connections unchanged; when n 1＝ 0 and n2 + n3 ≥ n, keep the existing connections unchanged and establish new connections; when n 1＝ 0 and n2 + n3 < n, disconnect all connections and wait for requests from other nodes to establish links; 2.4) Repeat steps 2.1) to 2.3) to cluster the dynamic network in real time according to the trajectory information of the nodes.

2. The multi-hop clustering method for improving the topology duration of an aeronautical ad hoc network according to claim 1 is characterized in that: in step 1.1), the method of preprocessing the original data of the flight trajectory of the aviation node to obtain the preprocessed trajectory attributes is: The original data of the flight trajectory of the aviation node is converted from the WGS-84 coordinate system to the Earth-centered Earth-fixed coordinate system. The conversion formula is shown in formula (1): Where B, L and H represent the latitude, longitude and altitude in the WGS-84 coordinate system respectively; x, y, z represent the ECEF coordinate position respectively; a represents the semi-major axis of the earth, which is 6,378,137m; b represents the semi-minor axis of the earth, which is 6,356,752m; N R represents the main vertical radius of curvature; e represents the first eccentricity of the ellipsoid.

3. The multi-hop clustering method for improving the topology duration of an aeronautical ad hoc network according to claim 1, characterized in that: In step 1.2), the method in which each node sends the pre-processed trajectory attributes to neighboring nodes in a broadcasting manner is: The preprocessed trajectory attributes are placed in a data packet in a fixed format, and the current node is used as the source node, and the data packet is sent to all one-hop neighbor nodes by broadcasting.

4. The multi-hop clustering method for improving the topology duration of an aeronautical ad hoc network according to claim 1, characterized in that: In step 1.3), each node obtains the trajectory attributes of all its neighbor nodes, and the method for generating the neighbor attribute table is: When a node broadcasts its own attributes, it will receive trajectory attribute data packets from neighboring nodes. The node collects the trajectory attributes of all neighboring nodes to generate a neighbor attribute table.

5. The multi-hop clustering method for improving the topology duration of an aeronautical ad hoc network according to claim 1, characterized in that: In step 1.4), the method for calculating the link duration between a node and its neighbor based on the trajectory attributes of the neighbor node is: According to the location attribute and mobility attribute of the node, the duration of the link between nodes is calculated; Link duration refers to the duration that two nodes can communicate with each other. The calculation formula is shown in formula (2): Among them, Stime ij represents the duration of time that nodes i and j are within its transmission range; D ij Indicates the relative position of node i and node j; V ij represents the relative speed between node i and node j; R represents the communication radius between node i and node j.

6. The multi-hop clustering method for improving the topology duration of an aeronautical ad hoc network according to claim 1, characterized in that: In step 1.5), the method for counting the number m of links greater than the link duration threshold is: Counting the number of links greater than the link duration threshold refers to counting the number of neighbors in the neighbor attribute table that meet the condition for establishing a node-to-node link.

7. The multi-hop clustering method for improving the topology duration of an aeronautical ad hoc network according to claim 1, characterized in that: In step 1.6), when the number of links greater than the link duration threshold exceeds the set minimum connectivity value n, the node submits a link establishment request to a neighboring node to form a network clustering method as follows: When the node satisfies the condition that the number of neighbors m for establishing a node-to-node link is greater than the minimum connectivity n, it sends a connection request to its neighbor nodes to establish a node-to-node link; otherwise, the node waits for connection requests from other nodes.

8. The multi-hop clustering method for improving the topology duration of an aeronautical ad hoc network according to claim 1, characterized in that: In step 2.1), the method for updating the trajectory information of the neighbor node in the neighbor attribute table is: First, each node preprocesses its flight trajectory raw data to obtain the preprocessed trajectory attributes, then broadcasts the preprocessed trajectory attributes while receiving the trajectory attributes of its neighbors, and then updates the neighbor attribute table according to the received trajectory attributes; In step 2.2), the method for calculating the link duration of the existing link and the link duration between the new neighbor node and the neighbor node according to the trajectory attribute of the neighbor node in the neighbor attribute table is: According to the location attribute and mobility attribute of the node, the link duration between nodes is calculated using formula (2).

9. The multi-hop clustering method for improving the topology duration of an aeronautical ad hoc network according to claim 1, characterized in that: In step 2.4), the method of repeating steps 2.1) to step 2.3) to cluster the dynamic network in real time according to the node trajectory information is: After each time period, step 2.1), step 2.2) and step 2.3) are sequentially executed to form a real-time and effective network clustering, thereby improving the continuity of the network topology.

Images