CN-122028139-A - Dynamic self-adaptive clustering method, device, equipment, medium and product of underwater acoustic sensing network

CN122028139ACN 122028139 ACN122028139 ACN 122028139ACN-122028139-A

Abstract

The application discloses a dynamic self-adaptive clustering method, a device, equipment, a medium and a product of an underwater sound sensing network, and relates to the technical field of underwater sound communication and networks, wherein the method comprises the steps of acquiring multidimensional data of each survival node of a current round in the underwater sound sensing network, inputting the multidimensional data into a trained lightweight neural network prediction model to predict and evaluate the multi-target utility of each survival node, and obtaining a predicted value of the multi-target utility of each survival node; the method comprises the steps of dynamically screening a plurality of cluster head nodes from all surviving nodes by utilizing a predictive value of multi-objective utility and adopting a game theory, and distributing common nodes into corresponding clusters by adopting a nearest neighbor principle to form a stable cluster structure, wherein the common nodes are surviving nodes except the cluster head nodes in all surviving nodes, and each cluster comprises one cluster head node and a plurality of common nodes. The application improves the stability of the clustering effect.

Inventors

WEI YAN
CHEN ZI
YUAN YU
Nie Foyuan
QU FENGZHONG

Assignees

浙江大学

Dates

Publication Date: 20260512
Application Date: 20260413

Claims (10)

1. A dynamic self-adaptive clustering method of an underwater acoustic sensing network is characterized by comprising the following steps: Acquiring multidimensional data of each survival node of the current wheel in the underwater sound sensing network, wherein the multidimensional data comprises three-dimensional position information, residual energy states, euclidean distance to a water surface base station, neighbor survival node density in a communication range, data transmission rate and stability indexes based on historical expression; The multi-objective utility of each survival node is predicted and evaluated by inputting the multi-dimensional data to a trained lightweight neural network prediction model, so as to obtain a predicted value of the multi-objective utility of each survival node; Dynamically screening a plurality of cluster head nodes from all surviving nodes by utilizing the predicted value of the multi-objective utility and adopting a game theory, wherein the surviving nodes are nodes with residual energy larger than or equal to preset energy; And distributing the common nodes to the corresponding clusters by adopting a nearest neighbor principle to form a stable cluster structure, wherein the common nodes are survival nodes except for the cluster head nodes in each survival node, and each cluster comprises one cluster head node and a plurality of common nodes.
2. The underwater acoustic sensor network dynamic adaptive clustering method of claim 1, wherein the lightweight neural network prediction model comprises an input layer, two hidden layers and an output layer which are sequentially connected.
3. The method of claim 1, wherein the multi-objective utilities include energy efficiency, coverage quality, communication latency, and network connectivity.
4. The method for dynamically and adaptively clustering the underwater acoustic sensor network according to claim 1, wherein the method for dynamically screening a plurality of cluster head nodes from all surviving nodes by using a game theory by using the predicted value of the multi-objective utility is characterized by comprising the following steps: Based on the predicted value of the multi-target utility, a cluster head candidate set is obtained by adopting a cluster head selection mechanism designed by a game theory; optimizing the cluster head candidate set by adopting a simulated annealing strategy to obtain an optimized cluster head node set; And screening a plurality of cluster head nodes from the optimized cluster head node set by utilizing the minimum distance constraint.
5. The method for dynamically and adaptively clustering the underwater acoustic sensor network according to claim 4, wherein the cluster head candidate set is obtained by adopting a cluster head selection mechanism designed by a game theory based on the predicted value of the multi-objective utility, and the method specifically comprises the following steps: determining the weight of each target utility according to the clustering targets, and carrying out weighted summation to obtain the comprehensive utility value of each survival node; Defining each survival node in the underwater sound sensing network as a game participant, and calculating a benefit function value of each survival node as a cluster head and a benefit function value of each survival node as a common node based on the comprehensive utility value; calculating the profit difference between the profit value of each surviving node as the cluster head and the profit function value of each surviving node as the common node, and updating the selection probability of each surviving node as the cluster head based on the profit difference; and obtaining a cluster head candidate set based on the selection probability.
6. The method for dynamically adaptively clustering a hydroacoustic sensor network of claim 5, wherein the optimized cluster head node set comprises a plurality of initial cluster head nodes; the method for selecting the cluster head nodes from the optimized cluster head node set by utilizing the minimum distance constraint comprises the following steps: Sequencing all initial cluster head nodes according to the comprehensive utility value from high to low to obtain an ordered list, wherein the initial cluster head nodes in the ordered list are used as investigation nodes; initializing a screening set to obtain an initial screening set, wherein the initial screening set is an empty set; Adding the investigation node with the highest comprehensive utility value in the ordered list to the initial screening set to serve as a screening node of the screening set to obtain the screening set; traversing the residual inspection nodes, adding the residual inspection nodes with the distance to the screening nodes in the screening set being greater than or equal to the minimum distance threshold value to the screening set to obtain a final screening set, and taking the final screening set as an initial cluster head set; the number of the cluster heads in the initial cluster head set is complemented to obtain a final cluster head set, and the final cluster head set comprises a plurality of cluster head nodes.
7. The utility model provides a dynamic self-adaptation clustering device of underwater sound sensing network which characterized in that includes: The acquisition module is used for acquiring multidimensional data of each survival node of the current wheel in the underwater sound sensing network, wherein the multidimensional data comprises three-dimensional position information, residual energy states, euclidean distances to a water surface base station, neighbor survival node density in a communication range, data transmission rate and stability indexes based on historical expression; the utility prediction module is used for inputting the multidimensional data into the trained lightweight neural network prediction model to carry out prediction evaluation on the multi-target utility of each survival node to obtain a predicted value of the multi-target utility of each survival node; The cluster head node screening module is used for dynamically screening a plurality of cluster head nodes from all surviving nodes by utilizing the predicted value of the multi-target utility and adopting a game theory, wherein the surviving nodes are nodes with residual energy larger than or equal to preset energy; The self-adaptive clustering module is used for distributing common nodes into corresponding clusters by adopting a nearest neighbor principle to form a stable cluster structure, wherein the common nodes are survival nodes except for cluster head nodes in all survival nodes, and each cluster comprises one cluster head node and a plurality of common nodes.
8. A computer device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, characterized in that the processor executes the computer program to implement the method of dynamic adaptive clustering of an underwater sound sensor network according to any of the claims 1-6.
9. A computer readable storage medium having stored thereon a computer program, which when executed by a processor implements the dynamic adaptive clustering method of an underwater sound sensing network according to any of claims 1-6.
10. A computer program product comprising a computer program, characterized in that the computer program, when executed by a processor, implements the dynamic adaptive clustering method of an underwater sound sensor network as claimed in any one of claims 1-6.

Description

Dynamic self-adaptive clustering method, device, equipment, medium and product of underwater acoustic sensing network Technical Field The application relates to the technical field of underwater acoustic communication and networks, in particular to a dynamic self-adaptive clustering method, a device, equipment, a medium and a product of an underwater acoustic sensing network. Background As an important underwater information acquisition means, the underwater acoustic sensor network has increasingly outstanding problems of network performance and energy efficiency. The traditional underwater acoustic sensing network clustering method, such as a low-power consumption self-adaptive clustering layered protocol (low-ENERGY ADAPTIVE clustering hierarchy, LEACH) and an improved algorithm thereof, often faces the problems of unreasonable cluster head selection, unbalanced energy consumption, poor adaptability to dynamic underwater environments and the like in the aspects of network topology management and energy optimization. Especially when processing complex scenes such as node mobility, energy limitation, time-varying of underwater acoustic channels and the like, the existing clustering method often cannot realize optimal network performance, so that the network life cycle is short and the communication quality is poor. The self-adaptive clustering method is used as a network topology control technology, and can dynamically adjust the clustering strategy according to the network state and the node characteristics, so that the network energy efficiency and the communication reliability are improved. However, the related adaptive clustering method still has a certain limitation in terms of decision dimension and optimization mechanism, and it is difficult to comprehensively consider multi-objective optimization requirements such as energy balance, coverage quality, communication delay and the like. Specifically, the traditional method only considers single index energy or distance, lacks multi-dimensional comprehensive evaluation of network state, and meanwhile, the clustering strategy with fixed parameters is difficult to adapt to dynamic change of underwater environment, so that the clustering effect is unstable. Disclosure of Invention The application aims to provide a dynamic self-adaptive clustering method, device, equipment, medium and product of an underwater acoustic sensing network, which can realize multi-dimensional comprehensive evaluation of network states and improve the stability of clustering effects. In order to achieve the above object, the present application provides the following solutions: In a first aspect, the present application provides a dynamic adaptive clustering method for an underwater acoustic sensor network, including: Acquiring multidimensional data of each survival node of the current wheel in the underwater sound sensing network, wherein the multidimensional data comprises three-dimensional position information, residual energy states, euclidean distance to a water surface base station, neighbor survival node density in a communication range, data transmission rate and stability indexes based on historical expression; The multi-objective utility of each survival node is predicted and evaluated by inputting the multi-dimensional data to a trained lightweight neural network prediction model, so as to obtain a predicted value of the multi-objective utility of each survival node; Dynamically screening a plurality of cluster head nodes from all surviving nodes by utilizing the predicted value of the multi-objective utility and adopting a game theory, wherein the surviving nodes are nodes with residual energy larger than or equal to preset energy; And distributing the common nodes to the corresponding clusters by adopting a nearest neighbor principle to form a stable cluster structure, wherein the common nodes are survival nodes except for the cluster head nodes in each survival node, and each cluster comprises one cluster head node and a plurality of common nodes. In a second aspect, the present application provides a dynamic adaptive clustering device for an underwater acoustic sensor network, including: The acquisition module is used for acquiring multidimensional data of each survival node of the current wheel in the underwater sound sensing network, wherein the multidimensional data comprises three-dimensional position information, residual energy states, euclidean distances to a water surface base station, neighbor survival node density in a communication range, data transmission rate and stability indexes based on historical expression; the utility prediction module is used for inputting the multidimensional data into the trained lightweight neural network prediction model to carry out prediction evaluation on the multi-target utility of each survival node to obtain a predicted value of the multi-target utility of each survival node; The cluster head node screening module is us