Self-adaptive layering and clustering method for cross-domain communication network

The adaptive hierarchical clustering method improves cross-domain network performance by determining node weights and updating topology based on association information, addressing throughput and cost issues in heterogeneous and dynamic networks.

CN119071876AInactive Publication Date: 2024-12-03UNIV OF ELECTRONICS SCI & TECH OF CHINA
View PDF 2 Cites 0 Cited by

Patent Information

Application Number
CN202411268161.9
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-09-11
Publication Date
2024-12-03
Estimated Expiration
Not applicable · inactive patent

AI Technical Summary

Technical Problem

Existing clustering methods for cross-domain communication networks, such as minimum ID clustering or highest node degree clustering, fail to effectively manage networks with heterogeneous nodes and dynamic topologies, leading to reduced throughput, increased cluster head load, and high maintenance costs.

Method used

A cross-domain communication network adaptive hierarchical clustering method that determines node weights based on multiple association information to form effective clusters, updating the network topology for improved performance.

Benefits of technology

Enhances network throughput and reduces clustering costs by optimizing node clustering based on node weights, effectively managing cross-domain communication networks with heterogeneous and dynamic topologies.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119071876A_ABST
    Figure CN119071876A_ABST
Patent Text Reader

Abstract

The invention discloses a self-adaptive layering and clustering method for a cross-domain communication network. The method comprises the following steps: determining at least one to-be-clustered sub-network corresponding to a to-be-processed cross-domain communication network; the to-be-clustered sub-network comprises a plurality of network nodes; for at least one to-be-clustered sub-network, determining a target node weight corresponding to each network node according to multiple pieces of node association information corresponding to each network node in the to-be-clustered sub-network; determining at least one cluster corresponding to the to-be-clustered sub-network according to the target node weights corresponding to the plurality of network nodes; and updating the network topology of the to-be-processed cross-domain communication network based on the at least one cluster corresponding to the at least one to-be-clustered sub-network, so as to carry out communication according to the updated to-be-processed cross-domain communication network. According to the technical scheme, the effects of determining the node weight according to various node association information of the network nodes in a cross-domain communication scene and performing effective node clustering on the cross-domain communication network based on the node weight of the network nodes are realized.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of communication technology, and in particular to a cross-domain communication network adaptive hierarchical clustering method. Background Art

[0002] For cross-domain communication scenarios, the air communication nodes, surface communication nodes, and underwater communication nodes in the cross-domain communication network have the characteristics of being distributed on multiple platforms, cross-domain, and large in number. The cross-domain communication system network can change the topology structure by clustering nodes, thereby adjusting the network performance.

[0003] In the related art, clustering algorithms (such as the minimum ID clustering algorithm or the highest node degree clustering algorithm) are usually used to perform node clustering. However, the above-mentioned node clustering method is only applicable to the node clustering of a single domain node. When the above-mentioned node clustering method is applied to a cross-domain communication network, due to the network and node heterogeneity of the cross-domain communication network, as well as the performance limitations and reliability of the underwater network, the number of clusters in the cross-domain communication network may be small, resulting in a large cluster head load. As the number of cluster member nodes increases, the throughput of the cross-domain communication network decreases, and the network performance is greatly reduced. In addition, when the node dynamics are high, the topological structure changes caused by the change in node position make the communication link less stable, the cluster head update frequency increases, and the maintenance cost of the cross-domain communication network clustering increases. Summary of the invention

[0004] The present invention provides a cross-domain communication network adaptive hierarchical clustering method to achieve the effect of determining node weights according to multiple node association information of network nodes in a cross-domain communication scenario, and performing effective node clustering on the cross-domain communication network based on the node weights of the network nodes.

[0005] According to one aspect of the present invention, a cross-domain communication network adaptive hierarchical clustering method is provided, the method comprising:

[0006] Determine at least one sub-network to be clustered corresponding to the cross-domain communication network to be processed; wherein the cross-domain communication network to be processed is a network in which network nodes included communicate between different domains; the sub-network to be clustered includes multiple network nodes;

[0007] For at least one of the sub-networks to be clustered, determining a target node weight corresponding to each of the network nodes in the sub-network to be clustered according to a plurality of node association information corresponding to each of the network nodes in the sub-network to be clustered;

[0008] Determine at least one cluster corresponding to the sub-network to be clustered according to target node weights corresponding to the plurality of network nodes;

[0009] The network topology of the cross-domain communication network to be processed is updated based on at least one of the clusters corresponding to at least one of the sub-networks to be clustered, so as to communicate according to the updated cross-domain communication network to be processed.

[0010] The technical solution of the embodiment of the present invention determines at least one sub-network to be clustered corresponding to the cross-domain communication network to be processed; wherein the sub-network to be clustered includes multiple network nodes, and further, for at least one sub-network to be clustered, the target node weight corresponding to each network node is determined according to multiple node association information corresponding to each network node in the sub-network to be clustered; at least one cluster corresponding to the sub-network to be clustered is determined according to the target node weights corresponding to the multiple network nodes; based on at least one cluster corresponding to at least one sub-network to be clustered, the network topology of the cross-domain communication network to be processed is updated, so as to communicate according to the updated cross-domain communication network to be processed, thereby solving the problems of decreased throughput of the cross-domain communication network after clustering, decreased network performance and high network clustering cost existing in the related art, and realizing the effect of determining node weights according to multiple node association information of network nodes in the cross-domain communication scenario, and performing effective node clustering on the cross-domain communication network based on the node weights of the network nodes, thereby improving the practicality of node clustering in the cross-domain communication network, and realizing the effect of effectively managing the cross-domain communication network by clustering the network, thereby improving the network performance of the cross-domain communication network.

[0011] It should be understood that the contents described in this section are not intended to identify the key or important features of the embodiments of the present invention, nor are they intended to limit the scope of the present invention. Other features of the present invention will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0013] Figure 1 This is a flowchart of a cross-domain communication network adaptive hierarchical clustering method provided according to the first embodiment of the present invention;

[0014] Figure 2 This is a schematic diagram of the effect of a cross-domain communication network to be processed provided in Embodiment 1 of the present invention;

[0015] Figure 3 This is a flowchart of a cross-domain communication network adaptive hierarchical clustering method provided according to Embodiment 2 of the present invention. DETAILED DESCRIPTION

[0016] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings 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 should fall within the scope of protection of the present invention.

[0017] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0018] Embodiment 1

[0019] Figure 1 This is a flowchart of a cross-domain communication network adaptive hierarchical clustering method provided by the first embodiment of the present invention. This embodiment is applicable to the case of clustering network nodes included in the cross-domain communication network. The method can be executed by a cross-domain communication network adaptive hierarchical clustering device. The cross-domain communication network adaptive hierarchical clustering device can be implemented in the form of hardware and / or software. The cross-domain communication network adaptive hierarchical clustering device can be configured in a terminal and / or a server. Figure 1 As shown, the method includes:

[0020] S110: Determine at least one sub-network to be clustered corresponding to the cross-domain communication network to be processed; wherein the sub-network to be clustered includes a plurality of network nodes.

[0021] Among them, the cross-domain communication network to be processed may be a cross-domain communication network to be processed by hierarchical clustering. The cross-domain network to be processed is used to characterize that the nodes included are located in different communication domains. It can be understood that the cross-domain communication network refers to the exchange of data or communication between network nodes when the network nodes are located in different communication domains. The communication domain refers to at least one of the protocol, domain name and port number. In this embodiment, the cross-domain communication network to be processed may refer to a communication network in which network nodes are located in at least two of the air domain, the surface domain and the underwater domain, and network nodes in different domains can perform network communication and data transmission. Exemplarily, the cross-domain communication network to be processed may include air nodes, surface nodes and underwater nodes. The sub-network to be clustered may be a sub-network divided by the cross-domain communication network to be processed, and the sub-network is a sub-network to be clustered. In this embodiment, the sub-network to be clustered may include network nodes corresponding to at least two communication domains. It can be understood that the network node may be a workstation, a network user or a personal computer, or a server, a printer or other network-connected device. Connecting multiple network nodes with communication lines to form a certain geometric relationship can obtain the network topology. Exemplarily, assuming that the cross-domain communication network to be processed may include aerial nodes, surface nodes, and underwater nodes, it can be determined that the number of sub-networks to be clustered corresponding to the cross-domain communication network to be processed is 2, namely, sub-network A to be clustered and sub-network B to be clustered. Among them, sub-network A to be clustered includes aerial nodes, and sub-network B to be clustered includes surface nodes and underwater nodes.

[0022] In this embodiment, a network including at least two communication domains can be obtained, and the obtained network can be used as a cross-domain communication network to be processed. Further, in order to process the cross-domain communication network to be processed, the cross-domain communication network to be processed can be divided into at least one sub-network to be clustered according to the communication domains where the network nodes are located.

[0023] Optionally, determining at least one to-be-clustered sub-network corresponding to the to-be-processed cross-domain communication network includes: determining at least two node layers corresponding to the to-be-processed cross-domain communication network, and determining at least one to-be-clustered network based on the at least two node layers.

[0024] Among them, the node layer can be understood as a network composed of network nodes located in the same communication domain. The node layer includes at least one network node. In this embodiment, at least two node layers include an aerial node layer, a surface node layer, and an underwater node layer. The aerial node layer can be a network composed of aerial nodes. Wireless communication links are used for data transmission between aerial nodes. The surface node layer can be a network composed of surface nodes. Wireless communication links are used for data transmission between surface nodes. The underwater node layer can be a network composed of underwater nodes. Hydroacoustic communication links are used for data transmission between underwater nodes. Surface nodes can perform cross-medium data transmission with underwater nodes through hydroacoustic communication links. Surface nodes can perform cross-medium data transmission with aerial nodes through wireless communication links. When underwater nodes need to communicate with aerial nodes, they can be relayed through surface nodes. Surface nodes can receive acoustic wave signals emitted by underwater nodes, and convert the received acoustic wave signals into electromagnetic wave signals and send them to aerial nodes. In addition, communication between underwater nodes and aerial nodes can be achieved.

[0025] As an optional implementation of this embodiment, when obtaining the cross-domain communication network to be processed, the cross-domain communication network to be processed can be divided into at least two node layers according to the communication domains in which the network nodes included in the communication network to be processed are located. Further, at least two node layers can be merged according to the physical distance between different node layers, and the merged network including multiple network nodes is used as the sub-network to be clustered.

[0026] For example, Figure 2 1 is a schematic diagram of an effect of a cross-domain communication network to be processed according to an embodiment of the present invention. Figure 2 As shown, the cross-domain communication network 21 to be processed can be divided into three node layers, namely, the aerial node layer 211, the surface node layer 212 and the underwater node layer 213. Among them, the aerial node layer 211 includes 6 aerial nodes, and the aerial nodes use wireless communication links for data transmission; the surface node layer 212 includes 5 surface nodes, and the surface nodes use wireless communication links for data transmission; the underwater node layer 213 includes 5 underwater nodes, and the underwater nodes use hydroacoustic communication links for data transmission. The surface nodes can perform cross-medium data transmission with the underwater nodes through hydroacoustic communication links. The surface nodes can perform cross-medium data transmission with the aerial nodes through wireless communication links. Further, the aerial node layer 211 can be used as the sub-network A to be clustered, and the surface node layer 212 and the underwater node layer 213 can be used as the sub-network B to be clustered.

[0027] S120 . For at least one sub-network to be clustered, determine a target node weight corresponding to each network node according to a plurality of node association information corresponding to each network node in the sub-network to be clustered.

[0028] Among them, the node association information can be understood as information associated with the network node. Optionally, the node association information includes node connectivity, node location information, number of adjacent nodes, node location information of adjacent nodes, energy receiving power of adjacent nodes received by the network node, node type, node remaining energy, etc. The target node weight can be understood as a characteristic value that characterizes the weight of the network node in the network to which it belongs.

[0029] In this embodiment, when at least one sub-network to be clustered is obtained, for at least one sub-network to be clustered, in order to reduce the communication volume of the sub-network to be clustered, reduce the energy consumption of network nodes, and extend the life cycle of the sub-network to be clustered, the network nodes included in the sub-network to be clustered can be clustered. Further, in order to determine the cluster head node and cluster member nodes in the sub-network to be clustered, the target node weight corresponding to each network node can be determined according to the node association information corresponding to each network node in the sub-network to be clustered.

[0030] As an optional implementation of this embodiment, for multiple network nodes in the sub-network to be clustered, multiple node features corresponding to the network node can be determined according to multiple node association information corresponding to the network node. Furthermore, the multiple node features can be converted into target feature values ​​through weighted summation or averaging operations, and the target feature values ​​are used as target node weights corresponding to the network node.

[0031] S130. Determine at least one cluster corresponding to the sub-network to be clustered according to target node weights corresponding to a plurality of network nodes.

[0032] Among them, a cluster refers to a set of network nodes with some kind of association or similarity. A cluster can be regarded as a subnetwork or subgraph formed by multiple nodes connected. It can be understood that a cluster can include a cluster head node and at least one cluster member node. The cluster head node can be used for data forwarding and become a gathering and forwarding node for network data transmission. Cluster member nodes usually only have the authority to communicate with the cluster head node, and can transmit data to the cluster head node, which will complete further processing and forwarding. In general, network nodes in communication networks usually have limited energy. By clustering, network nodes can first send data to the associated cluster head node, and the cluster head node aggregates the data before forwarding it to the base station. In turn, the amount of communication can be significantly reduced, the node energy consumption can be reduced, and the life cycle of the entire network can be extended.

[0033] In this embodiment, when target node weights corresponding to multiple network nodes are obtained, cluster head nodes and cluster member nodes can be determined from the multiple network nodes according to the target node weights of the network nodes. Furthermore, at least one cluster can be generated according to the cluster head nodes and cluster member nodes.

[0034] Optionally, at least one cluster corresponding to the sub-network to be clustered is determined based on target node weights corresponding to multiple network nodes, including: for multiple network nodes, at least one adjacent node adjacent to the network node is determined in the sub-network to be clustered, and a target node weight corresponding to a minimum weight is determined from the target node weight corresponding to the network node and the target node weight corresponding to at least one adjacent node; the network node corresponding to the target node weight with the minimum weight is used as a cluster head node, at least one adjacent node adjacent to the cluster head node is used as a cluster member node of the cluster head node, and a cluster is generated based on the cluster head node and the cluster member nodes; when each network node in the sub-network to be clustered belongs to a corresponding cluster, at least one generated cluster is used as at least one cluster corresponding to the sub-network to be clustered.

[0035] Among them, the cluster head node can be a special node in the network, which can be used to manage and coordinate other nodes in its cluster. The cluster head node receives data from other nodes in the cluster and performs data fusion processing to reduce data transmission and improve data transmission efficiency. Cluster member nodes are all nodes in the network structure except the cluster head node. There is a clear subordinate relationship between cluster member nodes and cluster head nodes. Each cluster member node can usually only belong to one cluster, that is, it belongs to one cluster head node.

[0036] As an optional implementation of this embodiment, for multiple network nodes, at least one adjacent node adjacent to the network node can be determined from the sub-network to be clustered. Further, the target node weight with the minimum weight can be determined from the target node weight corresponding to the network node and the target node weight corresponding to at least one adjacent node. Further, the network node corresponding to the target node weight with the minimum weight can be used as the cluster head node, and at least one adjacent node adjacent to the cluster head node can be used as a cluster member node of the cluster head node. Further, a cluster can be generated based on the cluster head node and at least one cluster member node. Further, when it is detected that each network node in the sub-network to be clustered belongs to a corresponding cluster, at least one generated cluster can be used as at least one cluster corresponding to the sub-network to be clustered.

[0037] S140: Update the network topology of the cross-domain communication network to be processed based on at least one cluster corresponding to at least one sub-network to be clustered, so as to communicate according to the updated cross-domain communication network to be processed.

[0038] Among them, network topology refers to the physical layout of interconnecting various devices with transmission nodes, and refers to the specific physical or logical arrangement of objects that constitute the network. It can be understood that network topology can be used to characterize the data transmission relationship between nodes in the network.

[0039] In this embodiment, when at least one cluster corresponding to at least one sub-network to be clustered is obtained, the network topology of the cross-domain communication network to be processed can be updated based on the at least one cluster obtained, so that the data transmission relationship between each network node in the updated cross-domain communication network to be processed and the function to be performed by each network node are updated. Further, communication can be performed based on the updated cross-domain communication network to be processed.

[0040] The technical solution of the embodiment of the present invention determines at least one sub-network to be clustered corresponding to the cross-domain communication network to be processed; wherein the sub-network to be clustered includes multiple network nodes, and further, for at least one sub-network to be clustered, the target node weight corresponding to each network node is determined according to multiple node association information corresponding to each network node in the sub-network to be clustered; at least one cluster corresponding to the sub-network to be clustered is determined according to the target node weights corresponding to the multiple network nodes; based on at least one cluster corresponding to at least one sub-network to be clustered, the network topology of the cross-domain communication network to be processed is updated, so as to communicate according to the updated cross-domain communication network to be processed, thereby solving the problems of decreased throughput of the cross-domain communication network after clustering, decreased network performance and high network clustering cost existing in the related art, and realizing the effect of determining node weights according to multiple node association information of network nodes in the cross-domain communication scenario, and performing effective node clustering on the cross-domain communication network based on the node weights of the network nodes, thereby improving the practicality of node clustering in the cross-domain communication network, and realizing the effect of effectively managing the cross-domain communication network by clustering the network, thereby improving the network performance of the cross-domain communication network.

[0041] Embodiment 2

[0042] Figure 3 It is a flowchart of a cross-domain communication network adaptive hierarchical clustering method provided in the second embodiment of the present invention. On the basis of the above-mentioned embodiment, the target node weight corresponding to each network node is determined according to the multiple node association information corresponding to each network node in the sub-network to be clustered, including: determining the multiple node features corresponding to each network node according to the multiple node association information corresponding to each network node in the sub-network to be clustered; for each network node, determining the target node weight corresponding to the network node based on the multiple node features corresponding to the network node. The specific implementation method can be referred to the technical solution of this embodiment. Among them, the technical terms that are the same or similar to the above-mentioned embodiments are not repeated here.

[0043] like Figure 3 As shown, the method includes:

[0044] S210: Determine at least one to-be-clustered sub-network corresponding to the to-be-processed cross-domain communication network.

[0045] S220: For at least one sub-network to be clustered, determine multiple node features corresponding to each network node according to multiple node association information corresponding to each network node in the sub-network to be clustered.

[0046] The node characteristics can be understood as information characterizing the node attributes of the network nodes. Optionally, the node characteristics include node degree, node average distance, node movement characteristics, node processing characteristics, node residual energy, etc.

[0047] In this embodiment, for at least one sub-network to be clustered, multiple node association information corresponding to each network node in the sub-network to be clustered can be obtained. Furthermore, for each network node, multiple node features corresponding to the network node can be determined according to the multiple node association information corresponding to the network node. Different node features can be determined according to different node association information. The following describes the determination method of each node feature.

[0048] Optionally, the node association information includes node degree; the node feature includes node degree difference; and multiple node features corresponding to each network node are determined according to the multiple node association information corresponding to each network node in the sub-network to be clustered, including: for each network node, determining the difference between the node degree corresponding to the network node and the preset node degree, and taking the absolute value of the difference as the node degree difference corresponding to the network node.

[0049] The node degree may be the number of adjacent nodes directly connected to the network node. The preset node degree may be the expected number of adjacent nodes directly connected to the network node.

[0050] As an optional implementation of this embodiment, for each network node in the sub-network to be clustered, the node degree corresponding to the network node can be obtained. Further, the difference between the node degree and the preset node degree can be determined, and the absolute value of the difference is used as the node degree difference corresponding to the network node.

[0051] It should be noted that the advantage of determining the node degree difference based on the preset node degree is that it defines the ideal connectivity of the node, preventing the network node from becoming a network bottleneck due to overload. At the same time, the heterogeneity of the platform in the cross-domain communication scenario is taken into account, and the load that the platform can bear and the communication indicators of the node itself are taken into consideration, rather than simply averaging the network load, thereby improving the throughput and robustness of the network.

[0052] Exemplarily, the node degree difference may be determined based on the following formula:

[0053] D c =|d c -N|

[0054] Among them, D c represents the node degree difference corresponding to the network node c; d c Represents the node degree corresponding to the network node c; N represents the preset node degree.

[0055] Optionally, the node association information includes the number of adjacent nodes, node position information and node position information of at least one adjacent node adjacent to the network node; the node feature includes the average node distance; and multiple node features corresponding to each network node are determined according to the multiple node association information corresponding to each network node in the sub-network to be clustered, including: for each network node, determining the node relative distance and the node position information of the network node and the node position information of at least one adjacent node adjacent to the network node; and determining the ratio between the node relative distance and the number of adjacent nodes as the average node distance corresponding to the network node.

[0056] The node location information may be location information representing the node in the space where the node is located. Exemplarily, the node location information may be the three-dimensional coordinates of the network node. The neighboring node may be a node that is directly connected to the network node and can communicate. The number of neighboring nodes may be the number of neighboring nodes that are directly connected to the network node. The average node distance may be used to represent the average distance between the network node and its adjacent neighboring nodes.

[0057] As an optional implementation of this embodiment, for each network node, the number of adjacent nodes corresponding to the network node, node location information, and node location information of at least one adjacent node adjacent to the network node can be obtained. Further, for at least one adjacent node adjacent to the network node, the relative distance between the network node and the adjacent node can be determined based on the node location information of the network node and the node location information of the adjacent node. Further, at least one relative distance can be summed to obtain the node relative distance sum. Afterwards, the ratio between the node relative distance sum and the number of adjacent nodes can be determined, and the ratio can be used as the node average distance corresponding to the network node.

[0058] Exemplarily, the average node distance may be determined based on the following formula:

[0059]

[0060] Among them, P c represents the average node distance corresponding to the network node c; (X c ,Y c ,Z c ) represents the node location information corresponding to the network node c; (X n ,Y n ,Z n ) represents the node location information of the adjacent node n; N represents the number of adjacent nodes.

[0061] Optionally, the node association information includes the number of adjacent nodes and the energy receiving power received by the network node from at least one adjacent node adjacent to it, the energy receiving power including the energy receiving power corresponding to the current moment and the energy receiving power corresponding to the previous moment; the node characteristics include node mobility characteristics; and multiple node characteristics corresponding to each network node are determined according to the multiple node association information corresponding to each network node in the sub-network to be clustered, including: for each network node, determining the total relative speed corresponding to the network node according to the energy receiving power corresponding to at least one adjacent node adjacent to the network node at the current moment and the energy receiving power corresponding to the previous moment; determining the ratio between the total relative speed and the number of adjacent nodes as the node mobility characteristic corresponding to the network node.

[0062] The energy receiving power may be the energy emitted by the adjacent nodes received by the network node in a unit time. The current moment may refer to the moment when the node clustering is currently being performed. The node mobility feature may characterize the average mobility speed between the network node and its adjacent nodes.

[0063] As an optional implementation of this embodiment, for each network node, at least one adjacent node adjacent to the network node can be determined. Further, for at least one adjacent node, the ratio between the energy receiving power corresponding to the adjacent node at the current moment and the energy receiving power corresponding to the previous moment can be determined, and the ratio can be logarithmically processed to obtain the value to be superimposed corresponding to the adjacent node. Further, the value to be superimposed corresponding to at least one adjacent node can be summed, and the obtained sum can be used as the total relative speed corresponding to the network node. Afterwards, the ratio between the total relative speed and the number of adjacent nodes can be determined, and the ratio can be used as the node movement feature corresponding to the network node.

[0064] It should be noted that the advantage of using node mobility characteristics as one of the determining factors in the target node weight is that it takes into account the characteristics of platform group mobility that may exist in the context of cross-domain communication, which is more in line with the actual requirements of communication, avoids frequent changes of cluster heads caused by rapid node movement, and greatly improves the stability of the cluster.

[0065] Exemplarily, the node movement feature may be determined based on the following formula:

[0066]

[0067] Among them, Mc represents the node mobility feature corresponding to the network node c; N represents the number of adjacent nodes; Indicates the energy receiving power received by the network node c from the adjacent node n at the current moment; Indicates the energy receiving power received by the network node c from the adjacent node n at the last moment.

[0068] Optionally, the node association information includes the node type; the node feature information includes the node processing feature; and multiple node features corresponding to each network node are determined according to the multiple node association information corresponding to each network node in the sub-network to be clustered, including: for each network node, obtaining the node type corresponding to the network node, and determining the node processing parameters corresponding to the node type based on the preset type association information, and using the node processing parameters as the node processing features corresponding to the network node.

[0069] Among them, the node type can be understood as the node obtained by dividing the node according to the node's attributes, location and / or purpose. Optionally, the node type includes backbone nodes, aggregation nodes and sensor nodes. Backbone nodes refer to nodes that undertake the task of association connection and transmission in the network structure, and are the core connection points in the network. Aggregation nodes refer to nodes that are responsible for managing the data transmission of the entire network and acting as repeaters in the network, and are the hubs of data circulation in the network. Sensor nodes refer to sensor units in the network, and are nodes in the network that can realize data collection, data processing and / or wireless communication. Generally, the node processing capability of backbone nodes is stronger, the node processing capability of aggregation nodes is second, and the node processing capability of sensor nodes is weaker. The type association relationship is used to indicate the association relationship between node types and node processing parameters. The node processing parameter can be a predetermined parameter used to characterize the node processing capability. The larger the node processing parameter, the weaker the node processing capability. Optionally, the node processing parameter corresponding to the backbone node is less than the node processing parameter corresponding to the aggregation node, and the node processing parameter corresponding to the aggregation node is less than the node processing parameter corresponding to the sensor node. Exemplarily, the node processing parameter corresponding to the backbone node is 1, the node processing parameter corresponding to the aggregation node is 2, and the node processing parameter corresponding to the sensor node is 3.

[0070] As an optional implementation of this embodiment, at least one node type may be predetermined, and the node processing parameters corresponding to each node type may be determined. Furthermore, the at least one node type and the corresponding node processing parameters may be associated and stored to obtain a type association relationship. Furthermore, for each network node in the network to be clustered, the node type of the network node may be obtained. Furthermore, based on the node type, a query may be performed in the preset type association relationship to determine the node processing parameters corresponding to the node type. Furthermore, the node processing parameters may be used as the node processing features corresponding to the network node.

[0071] In this embodiment, the node characteristics may also include node residual energy. For each network node in the network to be clustered, the node residual energy of the network node may be obtained.

[0072] S230 . For each network node, determine a target node weight corresponding to the network node based on a plurality of node features corresponding to the network node.

[0073] In this embodiment, for each network node, when multiple node features corresponding to the network node are obtained, the multiple node features can be processed according to a preset node weight determination method to obtain a target node weight corresponding to the network node. The preset node weight determination method may include weighted summation or averaging.

[0074] Optionally, the target node weight corresponding to the network node is determined based on multiple node features, including: normalizing the multiple node features and separately determining the feature weight corresponding to each node feature; processing the multiple normalized node features and the feature weight corresponding to each node feature through a weighted sum operation to obtain the target node weight corresponding to the network node.

[0075] The feature weight may be a preset parameter used to characterize the importance of the node feature in the node clustering. The feature weight may be any value. It should be noted that the sum of multiple feature weights may be a preset parameter, for example, 1.

[0076] As an optional implementation of this embodiment, multiple node features correspond to different dimensions. In order to unify the dimensions and to determine the target node weight later, the multiple node features can be normalized to obtain multiple normalized node features, and the feature weight corresponding to each node feature can be determined respectively. Furthermore, for multiple normalized node features, the product between the normalized node features and the feature weights corresponding to them can be determined, and the product can be used as the parameter to be superimposed corresponding to the node features. Furthermore, the parameters to be superimposed corresponding to the multiple node features can be added, and the parameter obtained after the addition can be used as the target node weight corresponding to the network node.

[0077] Exemplarily, the target node weight may be determined based on the following formula:

[0078] W c =e1D c +e2P c +e3M c +e4T c +e5I c

[0079] Among them, W c represents the target node weight corresponding to the network node c; e1, e2, e3, e4, e5 represent feature weights, e1+e2+e3+e4+e5=1; D c represents the normalized node degree difference; P crepresents the normalized average distance of nodes; M c represents the normalized node movement feature; T c Represents the normalized node residual energy; I c Represents the normalized node processing features.

[0080] Optionally, determining a target node weight corresponding to the network node based on multiple node features includes: normalizing the multiple node features, averaging the normalized multiple node features, and using the average value as the target node weight corresponding to the network node.

[0081] As an optional implementation of this embodiment, for multiple network nodes, when multiple node features corresponding to the network nodes are obtained, the multiple node features can be normalized to obtain multiple normalized node features. Further, the multiple normalized node features can be summed to obtain the total node features. After that, the ratio between the total node features and the number of node features can be determined. Furthermore, the ratio can be used as the target node weight corresponding to the network node.

[0082] S240: Determine at least one cluster corresponding to the sub-network to be clustered according to target node weights corresponding to the plurality of network nodes.

[0083] S250: Update the network topology of the inter-domain communication network to be processed based on at least one cluster corresponding to at least one sub-network to be clustered, so as to communicate according to the updated inter-domain communication network to be processed.

[0084] The technical solution of the embodiment of the present invention determines at least one sub-network to be clustered corresponding to the cross-domain communication network to be processed; wherein the sub-network to be clustered includes multiple network nodes, and further, for at least one sub-network to be clustered, the target node weight corresponding to each network node is determined according to multiple node association information corresponding to each network node in the sub-network to be clustered; at least one cluster corresponding to the sub-network to be clustered is determined according to the target node weights corresponding to the multiple network nodes; based on at least one cluster corresponding to at least one sub-network to be clustered, the network topology of the cross-domain communication network to be processed is updated, so as to communicate according to the updated cross-domain communication network to be processed, thereby solving the problems of decreased throughput of the cross-domain communication network after clustering, decreased network performance and high network clustering cost existing in the related art, and realizing the effect of determining node weights according to multiple node association information of network nodes in the cross-domain communication scenario, and performing effective node clustering on the cross-domain communication network based on the node weights of the network nodes, thereby improving the practicality of node clustering in the cross-domain communication network, and realizing the effect of effectively managing the cross-domain communication network by clustering the network, thereby improving the network performance of the cross-domain communication network.

Claims

1. A cross-domain communication network adaptive hierarchical clustering method, characterized in that: include: Determine at least one sub-network to be clustered corresponding to the cross-domain communication network to be processed; wherein the cross-domain communication network to be processed is a network in which network nodes included communicate between different domains; the sub-network to be clustered includes multiple network nodes; For at least one of the sub-networks to be clustered, determining a target node weight corresponding to each of the network nodes in the sub-network to be clustered according to a plurality of node association information corresponding to each of the network nodes in the sub-network to be clustered; Determine at least one cluster corresponding to the sub-network to be clustered according to target node weights corresponding to the plurality of network nodes; The network topology of the cross-domain communication network to be processed is updated based on at least one of the clusters corresponding to at least one of the sub-networks to be clustered, so as to communicate according to the updated cross-domain communication network to be processed.

2. The cross-domain communication network adaptive hierarchical clustering method according to claim 1, characterized in that: The determining the target node weight corresponding to each of the network nodes according to the plurality of node association information corresponding to each of the network nodes in the sub-network to be clustered comprises: Determine a plurality of node features corresponding to each of the network nodes according to a plurality of node association information corresponding to each of the network nodes in the sub-network to be clustered; For each of the network nodes, a target node weight corresponding to the network node is determined based on a plurality of the node features corresponding to the network node.

3. The cross-domain communication network adaptive hierarchical clustering method according to claim 2, characterized in that: The determining the target node weight corresponding to the network node based on the plurality of node features corresponding to the network node comprises: Normalizing the plurality of node features, and respectively determining a feature weight corresponding to each of the node features; The plurality of normalized node features and the feature weight corresponding to each of the node features are processed by weighted sum operation to obtain a target node weight corresponding to the network node.

4. The cross-domain communication network adaptive hierarchical clustering method according to claim 2, characterized in that: The determining the target node weight corresponding to the network node based on the plurality of node features corresponding to the network node comprises: The plurality of node features are normalized, and the average of the plurality of normalized node features is calculated, and the obtained average value is used as the target node weight corresponding to the network node.

5. The cross-domain communication network adaptive hierarchical clustering method according to claim 2, characterized in that: The node association information includes node connectivity; the node characteristics include node degree difference; The determining of a plurality of node features corresponding to each of the network nodes according to a plurality of node association information corresponding to each of the network nodes in the sub-network to be clustered comprises: For each of the network nodes, the difference between the node connectivity corresponding to the network node and a preset node degree is determined, and the absolute value of the difference is used as the node degree difference corresponding to the network node.

6. The cross-domain communication network adaptive hierarchical clustering method according to claim 2, characterized in that: The node association information includes the number of adjacent nodes, node location information, and node location information of at least one adjacent node adjacent to the network node; the node feature includes an average node distance; and the determining of multiple node features corresponding to each of the network nodes in the subnetwork to be clustered according to the multiple node association information corresponding to each of the network nodes in the subnetwork to be clustered includes: For each of the network nodes, determining a sum of node relative distances according to the node position information of the network node and the node position information of at least one adjacent node adjacent to the network node; A ratio between the node relative distance and the number of adjacent nodes is determined as an average node distance corresponding to the network node.

7. The cross-domain communication network adaptive hierarchical clustering method according to claim 2, characterized in that: The node association information includes the number of adjacent nodes and the energy receiving power of at least one adjacent node received by the network node, wherein the energy receiving power includes the energy receiving power corresponding to the current moment and the energy receiving power corresponding to the previous moment; the node characteristics include node mobility characteristics; The determining of a plurality of node features corresponding to each of the network nodes according to a plurality of node association information corresponding to each of the network nodes in the sub-network to be clustered comprises: For each of the network nodes, determining a total relative speed corresponding to the network node according to an energy received power corresponding to at least one of the neighboring nodes adjacent to the network node at a current moment and an energy received power corresponding to a previous moment; A ratio between the total relative speed and the number of adjacent nodes is determined as a node movement feature corresponding to the network node.

8. The cross-domain communication network adaptive hierarchical clustering method according to claim 2, characterized in that: The node association information includes node type; the node feature information includes node processing features; The determining of a plurality of node features corresponding to each of the network nodes according to a plurality of node association information corresponding to each of the network nodes in the sub-network to be clustered comprises: For each of the network nodes, the node type corresponding to the network node is obtained, and the node processing parameters corresponding to the node type are determined based on preset type association information, and the node processing parameters are used as the node processing characteristics corresponding to the network node; wherein the type association relationship is used to indicate the association relationship between the node type and the node processing parameters.

9. The cross-domain communication network adaptive hierarchical clustering method according to claim 1, characterized in that: The determining of at least one to-be-clustered sub-network corresponding to the to-be-processed cross-domain communication network includes: At least two node layers corresponding to the cross-domain communication network to be processed are determined, and at least one network to be clustered is determined based on the at least two node layers; wherein the at least two node layers include an aerial node layer, a surface node layer, and an underwater node layer.

10. The cross-domain communication network adaptive hierarchical clustering method according to claim 1, characterized in that: The determining at least one cluster corresponding to the sub-network to be clustered according to the target node weights corresponding to the plurality of network nodes comprises: For the plurality of network nodes, determining at least one adjacent node adjacent to the network node in the sub-network to be clustered, and determining a target node weight with a minimum weight value from the target node weight corresponding to the network node and the target node weight corresponding to at least one of the adjacent nodes; The network node corresponding to the target node weight with the minimum weight value is used as a cluster head node, and at least one adjacent node adjacent to the cluster head node is used as a cluster member node of the cluster head node, and a cluster is generated based on the cluster head node and at least one of the cluster member nodes; wherein the cluster member nodes in the same cluster are subordinate to the cluster head node; In the case that each of the network nodes in the sub-network to be clustered belongs to a corresponding cluster, at least one of the generated clusters is used as at least one cluster corresponding to the sub-network to be clustered.

Citation Information

Patent Citations

  • Self-adaptive on-demand weighted clustering method based on perceptron

    CN105744566A

  • Clustering routing method for cross-medium heterogeneous unmanned cluster system

    CN111586785A