Channel establishment and allocation method and device for unmanned platform

Abstract

The application provides a channel establishment and distribution method and device for unmanned platforms, the method comprising: acquiring a core available frequency band, a plurality of user nodes, a convergence gateway, a node quantity of the plurality of user nodes, and a communication distance of each user node to the convergence gateway; dividing the core available frequency band into a plurality of independent channels with equal widths according to the node quantity, and calculating an expected channel capacity of each user node on each independent channel according to a plurality of communication distances by using a communication attenuation model; constructing a bipartite graph matching model according to the plurality of user nodes, the plurality of independent channels, and the plurality of expected channel capacities; determining a target matching scheme according to the bipartite graph matching model, the target matching being a matching scheme maximizing the minimum expected channel capacity of the plurality of user nodes, and allocating a unique independent channel to each user node according to the target matching scheme. The problems of low communication efficiency and low stability existing in the prior art channel distribution method are solved.

Description

Technical Field

[0001] This application relates to the field of underwater communication technology, and more specifically, to a method and apparatus for establishing and allocating channels for unmanned platforms. Background Technology

[0002] Underwater communication is a core supporting technology for marine resource exploration and underwater environmental monitoring. Underwater communication differs from radio communication. Sound waves propagate at a very low speed in water, and the core frequency bands available for underwater communication are extremely limited. There is not enough spectrum for users to use at will. If channels are not allocated, multiple users will not compete for frequency bands, resulting in wasted frequency resources. Furthermore, if channels are not allocated, multiple users will transmit signals simultaneously on the same frequency band, causing serious co-channel conflicts and signal superposition interference.

[0003] Existing methods typically use fixed channel allocation, but the distance between different users and the gateway may be the same or different. Furthermore, the nodes are battery powered and are susceptible to displacement due to water flow. Therefore, existing fixed channel allocation methods result in poor efficiency and low fairness, as well as low spectrum utilization, which affects the efficient and stable communication between users and the gateway.

[0004] There is currently no effective solution to the problems of low communication efficiency and stability in existing channel allocation methods. Summary of the Invention

[0005] This application provides a method and apparatus for establishing and allocating channels for unmanned platforms, in order to solve the problems of low communication efficiency and low stability in existing channel allocation methods.

[0006] According to one aspect of the embodiments of this application, a channel establishment and allocation method for unmanned platforms is provided, comprising: obtaining a core available frequency band, multiple user nodes, a convergence gateway, the number of nodes of the multiple user nodes, and the communication distance from each user node to the convergence gateway, wherein the multiple user nodes and the convergence gateway are unmanned platform nodes; dividing the core available frequency band into multiple independent channels of equal width according to the number of nodes, and calculating the expected channel capacity of each user node on each independent channel based on the multiple communication distances using a communication attenuation model; constructing a bipartite graph matching model based on the multiple user nodes, multiple independent channels, and multiple expected channel capacities, wherein the first vertex set of the bipartite graph matching model is the user node set, the second vertex set is the independent channel set, and the edge weight between the first vertex and the second vertex is the expected channel capacity of the corresponding user node on the corresponding independent channel; determining a target matching scheme based on the bipartite graph matching model, wherein the target matching is a matching scheme that maximizes the minimum expected channel capacity of the multiple user nodes, and allocating a unique independent channel to each user node according to the target matching scheme.

[0007] According to another aspect of the embodiments of this application, a channel establishment and allocation apparatus for unmanned platforms is also provided, comprising: an acquisition unit, configured to acquire a core available frequency band, multiple user nodes, a convergence gateway, the number of nodes of the multiple user nodes, and the communication distance from each user node to the convergence gateway, wherein the multiple user nodes and the convergence gateway are unmanned platform nodes; a calculation unit, configured to divide the core available frequency band into multiple independent channels of equal width according to the number of nodes, and calculate the expected channel capacity of each user node on each independent channel based on the multiple communication distances using a communication attenuation model; a construction unit, configured to construct a bipartite graph matching model based on the multiple user nodes, multiple independent channels, and multiple expected channel capacities, wherein the first vertex set of the bipartite graph matching model is the user node set, the second vertex set is the independent channel set, and the edge weight between the first vertex and the second vertex is the expected channel capacity of the corresponding user node on the corresponding independent channel; and a determination unit, configured to determine a target matching scheme based on the bipartite graph matching model, wherein the target matching is a matching scheme that maximizes the minimum expected channel capacity of the multiple user nodes, and allocates a unique independent channel to each user node according to the target matching scheme.

[0008] According to another aspect of the embodiments of this application, a computer-readable storage medium is provided, which stores computer instructions for causing a computer to perform the channel establishment and allocation method for unmanned platforms described above.

[0009] According to another aspect of the embodiments of this application, an electronic device is also provided, the electronic device including: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores a computer program executable by the at least one processor, the computer program being executed by the at least one processor to cause the computer program to perform the above-described channel establishment and allocation method for unmanned platforms.

[0010] The aforementioned channel establishment and allocation method for unmanned platforms solves the problems of low communication efficiency and low stability in existing channel allocation methods. Attached Figure Description

[0011] To more clearly illustrate the technical solutions in the specific embodiments of this application or the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0012] Figure 1 This is a flowchart of an optional channel establishment and allocation method for unmanned platforms according to an embodiment of the present invention;

[0013] Figure 2 This is a schematic diagram of an optional channel establishment and allocation method for unmanned platforms according to an embodiment of the present invention; Figure 3 This is a schematic diagram of an optional channel establishment and allocation device for unmanned platforms according to an embodiment of the present invention. Detailed Implementation

[0014] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort should fall within the scope of protection of the present application.

[0015] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such terms can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein.

[0016] Underwater acoustic communication, widely used in single-hop aggregation and clustered networking scenarios, differs from radio communication in many ways. First, sound waves propagate at very low speeds in water. Furthermore, signal attenuation due to receiver noise is highly dependent on distance and frequency. Therefore, when designing an underwater communication system, transmission frequency and bandwidth should be carefully considered and selected based on the expected communication distance. The underwater acoustic spectrum is extremely scarce, so the communication resources available to a single user largely depend on the number of active users, which in turn depends on time and location.

[0017] Furthermore, acoustic transducers are more energy-intensive than radio waves, necessitating transmission schemes that conserve as much energy as possible. These facts must be considered when designing any underwater network. A major problem with MAC schemes operating in the time domain is their limited efficiency and poor scalability due to the significant underwater propagation delay. For this reason, dynamic spectrum allocation methods become very attractive, as their efficiency is unaffected by long propagation delays. In fact, using dynamic spectrum allocation methods imposes a greater signal processing workload on coherent detection because underwater channels tend to be highly time-varying, producing significant Doppler shifts.

[0018] However, the robust and efficient algorithms being actively researched for compensating for channel effects make dynamic spectrum allocation an increasingly viable option. On the other hand, the main problems with using dynamic spectrum allocation in underwater communications are: firstly, fixed channel allocation schemes cannot adapt to the characteristics of underwater acoustic propagation, failing to consider the distance differences between users and the aggregation gateway to match the transmission pattern of "lower frequencies for long distances, higher frequencies for short distances," resulting in a severe mismatch between channel capacity and node demand, and extremely low spectrum utilization; secondly, the lack of a maximum-minimum fair allocation mechanism prevents the maximization of the minimum channel capacity for all users, leading to short-distance users crowding out resources and long-distance users having excessively compressed transmission capacity, resulting in extremely poor system fairness. In short, due to the strong interaction between communication performance and the specific frequency band used, fixed channel allocation schemes traditionally used in radio communications are expected to produce poor efficiency and low fairness. Furthermore, the traditional TDMA fixed time slot scheduling suffers from low energy efficiency and lacks a node mobility-aware reallocation mechanism, further limiting the performance of familiar communication systems. Based on these shortcomings, existing technologies cannot meet the high-efficiency, fair, and stable communication requirements of underwater multi-physics cross-domain collaborative networking. Therefore, there is an urgent need for a channel establishment and allocation method adapted to underwater acoustic characteristics, possessing dynamic fair allocation, adaptive time slot scheduling, and mobility reallocation capabilities, suitable for unmanned platforms. Thus, to address these issues, this study employs a channel establishment and (dynamic) allocation method. From a communication efficiency perspective, this method not only improves communication efficiency but also significantly reduces node energy consumption and lowers communication costs. Simultaneously, this type of system exhibits strong scalability, enabling flexible deployment of communication networks of different sizes according to RW requirements, ensuring reliable and effective communication support in various complex RW scenarios.

[0019] To address the aforementioned issues, this application provides a channel establishment and allocation method for unmanned platforms. This method is applied to a monotonic underwater aggregation communication scenario, where the aggregation gateway is located at the center of an underwater area, multiple underwater user nodes are randomly distributed within the underwater area, and all underwater user nodes transmit data to the aggregation gateway. As an optional implementation method, please refer to... Figure 1This document illustrates a flowchart of a channel establishment and allocation method for unmanned platforms according to an embodiment of this application. The method may include at least one of the following steps (S102 to S108): S102, obtaining a core available frequency band, multiple user nodes, a convergence gateway, the number of nodes among the multiple user nodes, and the communication distance from each user node to the convergence gateway, wherein the multiple user nodes and the convergence gateway are unmanned platform nodes; S104, dividing the core available frequency band into multiple independent channels of equal width according to the number of nodes, and calculating the expected channel capacity of each user node on each independent channel based on the multiple communication distances using a communication attenuation model; S106, constructing a bipartite graph matching model based on the multiple user nodes, multiple independent channels, and multiple expected channel capacities, wherein the first vertex set of the bipartite graph matching model is the user node set, the second vertex set is the independent channel set, and the edge weight between the first vertex and the second vertex is the expected channel capacity of the corresponding user node on the corresponding independent channel; S108, determining a target matching scheme based on the bipartite graph matching model, wherein the target matching is the matching scheme that maximizes the minimum expected channel capacity of the multiple user nodes, and allocating a unique independent channel to each user node according to the target matching scheme.

[0020] The core available frequency band in S102 above can be understood as a specific frequency range that, after spectrum planning and avoiding interference frequency bands, can be used for data transmission between user nodes and the aggregation gateway in underwater communication scenarios. Its selection needs to adapt to the propagation characteristics of underwater acoustic communication, take into account the transmission distance and bandwidth requirements, and serve as the basis for subsequent channel allocation. It provides allocable spectrum resources for all user nodes, avoids spectrum interference between different devices, and ensures orderly communication. For example, in combination with the actual application of underwater acoustic communication, 1kHz~50kHz is selected as the core available frequency band. This frequency band has relatively small attenuation when propagating underwater, which can meet the needs of short, medium and long underwater communication, and avoids high-frequency interference frequency bands caused by shipping, marine life, etc. The aforementioned user nodes can be understood as terminal devices in an underwater cluster network responsible for collecting and generating data to be transmitted (such as marine environmental parameters, equipment operating status, etc.). They are the initiators of data transmission, belong to a certain cluster, and communicate directly only with the aggregation gateway (cluster head) of that cluster. The congested nodes, as terminals for underwater data acquisition and transmission, transmit various types of collected data (such as temperature, salinity, pressure, equipment energy consumption, etc.) to the aggregation gateway through allocated channels, and are the basic building blocks of the underwater communication network. For example, familiar temperature sensor nodes, underwater pressure sensor nodes, and communication nodes carried by unmanned underwater vehicles (UUVs) are all user nodes in this application. Assuming that this application sets up 5 user nodes, denoted as Node1 to Node5, all deployed in shallow sea areas at depths of 10 to 50 meters, responsible for collecting marine environmental data from different locations. A convergence gateway can be understood as the core control device in an underwater cluster network, also functioning as a cluster head node. It is responsible for receiving data transmitted from all user nodes within its cluster, performing channel allocation and management, time slot scheduling, Doppler frequency shift detection, etc., and acts as a bridge between user nodes and external networks (such as surface buoys and shore-based centers). It aggregates spectrum resources and communication scheduling within its cluster, receives and forwards data from user nodes, monitors communication link status, and dynamically adjusts channel and time slot allocation schemes to ensure the stability and efficiency of the entire cluster network. For example, an intelligent convergence gateway deployed in an underwater area at a depth of 30 meters has functions such as underwater acoustic communication, data processing, and scheduling control. As a cluster head node, it manages the five user nodes (Nodes 1-5) mentioned above and can forward received marine environmental data to surface buoys, realizing cross-domain data transmission between underwater and surface environments.

[0021] The independent channels in S104 above can be understood as non-overlapping frequency sub-bands of equal width, formed by dividing the core available frequency band according to the number of user nodes. Each independent channel is allocated to only one user node, avoiding co-channel interference between different user nodes. Each independent channel provides a dedicated communication frequency band for each user node, ensuring that data transmission between the user node and the aggregation gateway does not interfere with each other, improving channel utilization and communication quality. For example, if the core available frequency band is 1kHz~6kHz (bandwidth 5kHz) and there are 5 user nodes, then this frequency band is divided into 5 independent channels of equal width, each with a bandwidth of 1kHz: Channel 1 (1kHz~2kHz), Channel 2 (2kHz~3kHz), Channel 3 (3kHz~4kHz), Channel 4 (4kHz~5kHz), and Channel 5 (5kHz~6kHz). The aforementioned communication attenuation model is a mathematical model used to describe the signal strength attenuation caused by factors such as medium absorption, scattering, and reflection during the propagation of underwater communication signals. Its core function is to calculate signal loss at different communication distances. This model can quantify the impact of communication distance on signal transmission. By calculating the signal attenuation and combining it with noise power, the expected channel capacity of each user node on different channels can be accurately obtained, providing data support for channel allocation. For example, this application uses a "tone attenuation model + Thorp formula" to construct the communication attenuation model. The Thorp formula is used to calculate the seawater absorption loss during underwater acoustic propagation, while the tone attenuation model is used to calculate the total propagation loss of the signal at a specific distance. The combination of these two methods can accurately reflect the attenuation law of underwater signals. The aforementioned expected channel capacity refers to the maximum stable data rate that can be transmitted between a user node and the aggregation gateway under specific communication distance and channel conditions. It is a core indicator for measuring channel communication capability and is determined by the total signal propagation loss and noise power spectrum. For example, if the communication distance between Node1 and the aggregation gateway is 800m, the expected channel capacity on channel 1 is 2.5kbps, which means that when Node1 communicates with the aggregation gateway through channel 1, the maximum stable data rate that can be transmitted is 2.5kbps. If the actual data transmission requirement of Node1 is 2kbps, then this channel can meet its requirements.

[0022] The bipartite graph matching model in S106 above can be understood as a graph theory model for solving the "user-resource" allocation problem. It consists of two disjoint vertex sets (user node set and independent channel set) and edges connecting the two vertex sets. The edge weight is the expected channel capacity of the corresponding user node on the corresponding independent channel. The expected channel capacity is used as the edge weight of the bipartite graph matching model to determine the degree of fit between different user nodes and different channels, providing a basis for determining the target matching scheme and ensuring that the allocated channel can meet the data needs of the user nodes. In this application, the channel allocation problem is transformed into a graph matching problem. By solving this model, the optimal matching of "user node-independent channel" is achieved, ensuring that the expected channel capacity of all user nodes is maximized, taking into account both fairness and efficiency. For example, Node1 to Node5 are the first vertex set (user node set), and Channel1 to Channel5 are the second vertex set (independent channel set). The weight of each edge is the expected channel capacity of the corresponding user node on that channel (e.g., the weight of Node1-Channel1 is 2.5kbps, and the weight of Node1-Channel2 is 2.3kbps). A bipartite graph matching model is constructed to solve for the optimal channel allocation scheme.

[0023] The following examples illustrate S102 to S108 above, for example: S102, Acquire Core Parameters: Collect all core parameters required for channel establishment and allocation, ensuring the accuracy and completeness of the parameters. Among them, the core available frequency band can be selected according to the underwater communication scenario and interference situation, the number of user nodes can be deployed according to monitoring needs, and the communication distance can be measured through underwater positioning technology (such as underwater acoustic positioning). For example, the acquired core parameters include: the core available frequency band is 1kHz~6kHz (bandwidth 5kHz); the number of user nodes is 5 (Node1~Node5); the aggregation gateway is GW1 (cluster head node); and the communication distances from each user node to GW1 are Node1d=800m, Node2d=1000m, Node3d=1200m, Node4d=1500m, and Node5d=1800m, respectively.

[0024] S104, Divide independent channels and calculate expected channel capacity: Divide the core available frequency band into independent channels of equal width according to the number of user nodes (ensuring that the number of channels is consistent with the number of user nodes, and each user node can be allocated an independent channel). Simultaneously, using a communication attenuation model and considering the communication distance of each user node, calculate the expected channel capacity of each user node on each independent channel, quantifying the channel adaptability. For example, if there are 5 user nodes, the core available frequency band from 1kHz to 6kHz is divided into 5 independent channels of equal width, each with a bandwidth of 1kHz (Channel 1: 1kHz~2kHz, Channel 2: 2kHz~3kHz, Channel 3: 3kHz~4kHz, Channel 4: 4kHz~5kHz, Channel 5: 5kHz~6kHz). The expected channel capacity (unit: kbps) of each user node on different channels is calculated using the communication attenuation model, as shown in Table 1 below:

[0025] Table 1 S106, Construct a bipartite graph matching model: Using the user node set as the first node set and the independent channel set as the second node set, the channel capacity of each user node on the corresponding channel is used as the weight between two vertices to construct a bipartite graph matching model. This model is used to find the optimal "user-channel" allocation relationship through graph theory matching. For example, the first vertex set is {Node1, Node2, Node3, Node4, Node5}, the second vertex set is {Channel1, Channel2, Channel3, Channel4, Channel5}, and the edge weights are the values ​​in Table 1 above, such as the edge weight of Node1-Channel1 is 2.5, the edge weight of Node1-Channel2 is 1.3, and so on, to construct a complete bipartite graph matching model.

[0026] S108, based on a bipartite graph matching model, uses iterative deletion operations to find a target matching scheme that maximizes the minimum expected channel capacity of multiple user nodes (preferring to reduce the capacity of nearby user nodes while raising the minimum capacity of distant user nodes to ensure that all nodes in the network can communicate normally and that no node has a capacity of 0 and communication is interrupted). The target matching scheme ensures that each user node is allocated a unique independent channel, and that the minimum expected channel capacity of all user nodes is maximized, balancing fairness and practicality. For example, using the bipartite graph matching model, the final target matching scheme is: Node1 → Channel 1 (2.5kbps), Node2 → Channel 2 (2.0kbps), Node3 → Channel 3 (1.5kbps), Node4 → Channel 4 (1.0kbps), Node5 → Channel 5 (0.5kbps). At this point, the minimum expected channel capacity of all user nodes is 0.5kbps, which has been maximized. Subsequently, each user node is allocated a unique independent channel according to this scheme.

[0027] By adopting the above-described embodiments in this application, the problems of low communication efficiency and low stability in existing channel allocation methods are solved, and the efficiency and stability of underwater communication are improved.

[0028] The above-mentioned communication attenuation model calculates the expected channel capacity of each user node on each independent channel based on multiple communication distances, including: S1, establishing a communication attenuation model based on the tone attenuation model and Thorp's formula, wherein Thorp's formula is used to calculate the seawater absorption loss of underwater acoustic propagation; S2, using the tone attenuation model, calculating the total propagation loss after monotonic audio propagation at each communication distance based on multiple communication distances and seawater absorption loss; S3, calculating the noise power spectrum based on the acquired turbulence noise component, shipping noise component, wind and wave noise component, and thermal noise component; S4, calculating the expected channel capacity using the channel capacity theorem based on the noise power spectrum and the total propagation loss.

[0029] The above pitch decay model can be understood, but is not limited to, the Urick pitch decay model (i.e., based on Robert...). J Named after Robert J. Urick, this is a mathematical model used to calculate the propagation loss of underwater sound waves. Sound waves travel at the speed of sound in seawater. Propagation speed is affected by depth, temperature, and salinity. If depth, temperature, and salinity are constant, then the most important characteristic of underwater acoustic communication is the correlation between the optimal transmission frequency and bandwidth and the distance between communication nodes. In this application, communication frequency is used as the key factor. The transmission pitch attenuation model can be expressed as: ,in, It is the distance between the user node (transmitter) and the aggregation gateway (receiver). It corresponds to the path loss coefficient in terrestrial wireless circuits, used to model the geometry of propagation, and usually uses actual values. The factors in the above formula To account for seawater absorption loss (measured in dB / km, also known as attenuation), this application also models the conversion of sound pressure into heat. (Coefficient) It can be approximated by Thorp's formula: ,in, For frequencies in kHz The above equation returns the result in dB / km. .attenuation As the frequency changes significantly, the following patterns emerge: Low frequency: absorption is minimal, attenuation is mainly dominated by spread loss, and the signal can travel very far, making it suitable for distant user nodes; Mid frequency: absorption increases approximately linearly with frequency, and high-frequency signals attenuate faster; High frequency: absorption increases sharply, limiting long-distance transmission, and attenuation is extremely rapid, making it suitable for short-distance user nodes. Furthermore, the exponential term in the above formula... The presence of this enhances the distance dependence of the decay.

[0030] The above S1 can be understood as follows: Combining underwater acoustic propagation characteristics, a communication attenuation model is established based on the pitch attenuation model and Thorp's formula. Thorp's formula calculates absorption loss based on frequency, seawater temperature, etc., while the pitch attenuation model integrates factors such as propagation distance and seawater loss to calculate the total propagation loss. Assuming the seawater temperature is 25 degrees Celsius and the communication frequency is 1 kHz (channel 1), the calculation using the above Thorp formula yields: dB / km, combined with the tone attenuation model, integrates the diffusion loss caused by the propagation distance to obtain the total propagation loss model (i.e., the communication attenuation model).

[0031] S2 can be understood as follows: Using the established tone attenuation model, combining the communication distance of each user node and seawater absorption loss, the total propagation loss after monotonic audio propagation at each communication distance is calculated; the total propagation loss includes seawater absorption loss, diffusion loss, scattering loss, etc. The noise power spectrum in S3 is used to measure the channel noise level. The noise power spectrum (noise power spectral density PSD) is also frequency-dependent and is usually expressed as: It comprises the superposition of four contributions: turbulence, shipping and other human activities, wind and waves, and thermal noise in the receiver circuitry.

[0032] Turbulent noise components:

[0033] Ship noise components:

[0034] Surface wave noise components:

[0035] Thermal noise components:

[0036] in, represent , represent , , , . In It is a shipping factor, which represents the intensity of shipping activities on the water, and its value is between 0 and 1. Factors in The wind speed is measured in m / s. Different components affect the noise PSD at different frequencies. For example, in the high part of the sound spectrum, which is typically used for short-distance (tens of meters) transmission, the turbulence and shipping components have a smaller impact, while the other two components may dominate. In S4, the expected channel capacity of each user node on each independent channel is calculated based on the channel capacity theorem (Shannon formula), combined with the noise power spectrum and total propagation loss.

[0037] Understandably, after calculating the noise power spectral density, this application also calculates the signal-to-noise ratio to understand the impact of frequency-dependent channel effects on allocation measurements, thereby assisting in channel allocation based on frequency. Send and propagate distance The average SNR (Signal-to-Noise Ratio) of a tone is defined as: ;in, It's the transmission power. It is the noise power spectral density (i.e., the noise power spectrum mentioned above) (assuming that in Nearby narrow band (where is a constant), in the above formula, It is a frequency-dependent term. It increases with increasing frequency, while Then it decreases with increasing frequency; therefore, these two factors ( and The reciprocal of the product of ) for a certain frequency It has a maximum value (at frequency) In this context, the transmitted tone produces the most powerful propagation / noise conditions based on distance. This maximum value represents the optimal frequency for transmitting the tone.

[0038] The above S4, assuming as follows Figure 2 As shown, there are two user nodes (e.g. Figure 2User node 1 and user node 2 (as shown) share a common bandwidth (i.e., the aforementioned core available frequency band) to transmit to the destination (i.e., the aforementioned aggregation gateway). User node 1 is located 1 kilometer away, while user node 2 is located even further, 10 kilometers away from the destination. The system bandwidth ranges from 10 kHz to 40 kHz, divided into two equally wide channels, each 15 kHz. In this case, this application may result in two allocation outcomes: First, user node 1 is assigned to channel 1 (i.e., channel 1 is allocated to user node 1, i.e., as shown in the example). Figure 2 A to B (shown as the final allocation result of channel 1) and user node 2 on channel 2 (i.e., channel 2 is allocated to user node 2, i.e., as shown) Figure 2 E to F shown represent the final allocation result for channel 2); or the second option: user node 2 on channel 1 (i.e., channel 1 is allocated to user node 2, i.e., as shown in the diagram). Figure 2 As shown in the diagram, D to E represent the final allocation result of channel 1, and user node 1 on channel 2 (i.e., channel 2 is allocated to user node 2, as shown in the diagram). Figure 2 (B to C shown represent the final channel 1 allocation results). Depending on the chosen allocation, each user will experience a different channel response. Assuming both user nodes use the same power transmission, user node 2 will have the lowest throughput because of the greater distance and worse channel impact. Furthermore, assuming the goal is to maximize the user experience to achieve the minimum throughput, user 2's throughput must be maximized. In this case, the first allocation result is not optimal because the combined attenuation and noise effects are very large and severely limit throughput. Conversely, the second allocation result is optimal; that is, the optimal allocation scheme in this application (i.e., the target matching scheme described above) allocates the channel with the disclosure frequency to the farther user, as this typically corresponds to a stronger propagation effect. Based on the above characteristics, this application calculates the channel capacity for a specific underwater acoustic communication scenario in the following manner, assuming... and These are the upper and lower frequencies of a specific channel used for communication, and it is also assumed that the signal to be transmitted has a spectrum. According to Shannon-Hartley's theorem, channel capacity can be calculated using the following formula: This formula can be applied to any signal or noise frequency. In the following description, it is assumed that the user's throughput on a given channel is equal to the channel capacity calculated using the above formula.

[0039] Through the above-described embodiments of this application, the expected channel capacity of each user node on each independent channel can be calculated, thereby facilitating the provision of an accurate basis for the channel allocation described below.

[0040] The above-mentioned determination of the target matching scheme based on the bipartite graph matching model includes performing an iterative deletion operation based on the bipartite graph matching model and determining the target matching scheme based on the iterative deletion result. The iterative deletion operation includes: S1, determining multiple vertex edges between the first vertex set and the second vertex set corresponding to the bipartite graph matching model as a candidate edge set, and determining the candidate edge corresponding to the smallest edge weight as the smallest weight edge according to the edge weight corresponding to each candidate edge; S2, removing the smallest weight edge from the candidate edge set to obtain a reference candidate edge set, and determining whether the reference candidate edge set meets the reference matching requirement, which is used to indicate the matching that can allocate different independent channels to each user node; S3, if the reference matching requirement is met, the reference candidate edge set is determined as the candidate edge set and the iterative deletion operation continues; if the reference matching requirement is not met, the smallest weight edge is determined as the locked allocation edge, and the weight value of the locked allocation edge is determined as the allocation capacity threshold; S4, removing candidate edges corresponding to the same user node or the same channel as the locked allocation edge from the candidate edge set. When the number of locked allocation edges is equal to the number of user nodes, the iterative deletion operation terminates, thereby determining the target matching scheme. In addition, after determining the locked allocation edge, this application also includes: removing candidate edges from the candidate edge set whose edge weights are lower than the weights corresponding to the locked allocation edge.

[0041] Iterative deletion is the core operation for solving bipartite graph matching models. It involves repeatedly deleting candidate edges with the smallest weights, determining whether the remaining candidate edges meet the matching requirements, gradually locking in the optimal matching edge, and finally obtaining the target matching scheme. This process aims to achieve the matching objective of "maximizing the minimum expected channel capacity," ensuring that the channels allocated to all user nodes have the largest possible minimum expected channel capacity, preventing some user nodes from being unable to meet their transmission needs due to being allocated low-capacity channels. For example, for a bipartite graph, the edge with the smallest weight (e.g., Node5 - Channel 5, weight 1.2kbps) is first deleted. It is then determined whether the remaining edges can allocate different channels to all user nodes. If so, the edge with the smallest weight among the remaining edges is deleted; if not, this edge with the smallest weight is locked as the allocation edge, and this process is repeated until the allocation edges for all user nodes are locked.

[0042] S1 can be understood as: determining the candidate edge set and the minimum weight edge, that is, taking all edges connecting user nodes and independent channels in the bipartite graph matching model as the candidate edge set, traversing the weights of all candidate edges, and finding the edge with the smallest weight (i.e., the minimum weight edge); the channel capacity corresponding to the minimum weight edge is the smallest, and it is the object to be deleted first. For example, the candidate edge set contains all user nodes and all channels, a total of 25 edges; traversing all edge weights, the minimum weight edge is found to be Node5-channel 5 (weight 0.5kbps).

[0043] S2 involves deleting the edge with the lowest weight and determining the reference matching requirement. This means deleting the edge with the lowest weight from the candidate edge set to obtain a reference candidate edge set. It then checks whether the reference candidate edge set meets the "reference matching requirement" (i.e., whether different independent channels can be allocated to each user node to achieve a complete match). If it does, the deletion continues iteratively. If not, the edge with the lowest weight is locked as the allocation edge. For example, after deleting Node5-channel 5, the reference candidate edge set has 24 edges remaining. It then checks whether a complete match can be achieved: Node5 can choose channels 1-4 (weights 1.3, 1.1, 0.9, 0.7), and other user nodes can also choose the remaining channels. Therefore, the reference matching requirement is met, and the deletion continues iteratively.

[0044] In step S3 above, determining whether the candidate edge set contains a perfect match (i.e., whether it meets the reference matching requirement) can be achieved using the Hungarian algorithm to perform a maximum matching search in the bipartite graph. If the number of edges in the maximum matching is equal to the total number of user nodes, then a perfect match is determined to exist. Step S3 involves iterative deletion and locking of allocated edges. This involves repeating steps S1 to S2 above, each time deleting the smallest weight edge from the current candidate edge set and determining whether it meets the reference matching requirement. If it does not meet the requirement, the smallest weight edge is locked as an allocated edge, its weight is recorded as the allocation capacity threshold, and all candidate edges corresponding to the locked edge and the same user node or the same channel are deleted. The iteration continues until the number of locked allocated edges is equal to the number of user nodes, at which point the iteration terminates, and the locked allocated edges become the target matching scheme. For example: Continue deleting the edge with the lowest weight (Node4-Channel 5, weight 0.8kbps) from the remaining edges, and determine if the reference candidate edge set still meets the matching requirements; continue deleting Node5-Channel 4 (0.7kbps), Node4-Channel 4 (1.0kbps), Node5-Channel 3 (0.9kbps), Node3-Channel 5 (1.1kbps), Node4-Channel 3 (1.2kbps), Node5-Channel 2 (1.1kbps), and Node3-Channel 4 (1.3kbps). The nodes are Node2-Channel 5 (1.4kbps), Node4-Channel 2 (1.4kbps), and Node3-Channel 3 (1.5kbps). If Node5-Channel 1 (1.3kbps) is deleted, then Node5 will have no available channel and will not meet the reference matching requirements. Therefore, Node5-Channel 1 (1.3kbps) is locked as the allocation edge, and the allocation capacity threshold is 1.3kbps. All candidate edges related to Node5 and Channel 1 are deleted, and the iteration continues until 5 allocation edges are locked to obtain the target matching scheme.

[0045] The removal of candidate edges with weights lower than those corresponding to the locked and allocated edge can be understood as a supplementary operation after locking and allocating the edge. Specifically, after determining the locked and allocated edge, in addition to deleting candidate edges corresponding to the same user node or channel as the locked edge, all candidate edges with weights lower than the weight of the locked and allocated edge must also be deleted. The purpose is to ensure that the weights of the remaining candidate edges are not lower than the current allocated capacity threshold, further improving the channel capacity level for subsequent matching. For example, after locking Node5-Channel1 (weight 1.3kbps), all candidate edges with weights lower than 1.3kbps are deleted (such as Node4-Channel2 (1.4kbps, not deleted), Node3-Channel3 (1.5kbps, not deleted), Node2-Channel4 (1.6kbps, not deleted), Node4-Channel3 (1.2kbps, deleted), Node3-Channel4 (1.3kbps, not deleted), etc.), ensuring that the weights of the remaining candidate edges are all ≥1.3kbps, providing higher-quality candidate edges for subsequent iterations.

[0046] In determining the target matching scheme, this application considers a scenario where the available spectrum is divided into channels of equal width. Each user will be assigned a single channel (i.e., an independent channel), and each channel will be allocated to a single user. In principle, it is feasible to associate all user nodes with channels, but user nodes may be located in different locations, and attenuation and noise vary with frequency and distance. Therefore, in this application, each user node will obtain a different communication capacity (i.e., expected channel capacity) for each channel it is assigned. From the perspective of each user node, the optimal channel for a given user node is the channel with the highest capacity for that user node: high-frequency channels are more suitable for nearby user nodes, while low-frequency channels are better for distant user nodes. However, due to the special shape of the attenuation noise function, it is usually impossible to allocate its optimal channel to each user, so a trade-off must be made. In practical scenarios with limited transmission power, even using suboptimal channels, nearby user nodes will obtain fairly high capacity, while distant user nodes may experience very low capacity unless they are allocated a sufficiently low-frequency channel. In this regard, a good channel allocation scheme will be able to enhance the capacity of distant users without sacrificing too much capacity for nearby user nodes.

[0047] To achieve the maximum-minimum fair channel allocation scheme, for any feasible channel allocation Channel capacity Each subscript indicates a possible assignment. Capacity managed on a specific channel. For all feasible allocations. ,make according to Sort, for any other Feasible allocation They all Then feasible channel allocation This is the maximum-minimum fair allocation (i.e., the target allocation scheme mentioned above).

[0048] The channel allocation method in this application is as follows: Let... The number of users is equal to the number of channels to be allocated. In this application, the channel allocation problem is modeled as a matching problem on a bipartite graph (i.e., the bipartite graph matching model mentioned above). Let the vertices be... Representing the user, and using vertices This represents a channel. The assigned solution is an edge. set , making ,and It is a match. However, existing exhaustive search algorithms are not suitable for finding the maximum-minimum capacity allocation because the number of feasible allocations is... ,set up It is the set of all edges that can belong to a solution, i.e. This application is made by continuously from We work by removing edges with the lowest capacity until no better solution exists. As an edge, The capacity is the maximum of the minimum channel capacity in each of all feasible allocations; therefore, The capacity will only appear in the capacity vector of the maximum-minimum fair channel allocation. To find a complete maximum-minimum fair solution (i.e., all channel users are allocated), by removing... or All edges appearing in (because of the user) and channel (already assigned) to repeat the algorithm, when Users were assigned to The algorithm terminates when the problem on each channel is resolved, i.e., all users and channels have been allocated. This application implements the classic max-min fair allocation algorithm in the underwater acoustic dynamic spectrum allocation method. The pseudocode of the specific max-min fair channel allocation algorithm used in this application is shown below: Po:=P; Qo:=Ø; k:=0; while |Qk| <N do increment k (i,j):=MinCapacityEdge(Pk-1) Pk:=Pk-1\{(i,j)} Mk:=HighestCardinalityMatching(Pk) if |Mk|+|Qk-1| <N then Qk:=Qk-1U(i,j) for all m such that(i,m)∈Pk do Pk:=P\{(i,m)} end for for all n such that (n,j)∈Pk do Pk:=Pk\{(n,j)} end for else Qk:=Qk-1 end if end while return Qk The target matching scheme can be quickly determined through the above-described embodiments in this application.

[0049] After determining the target allocation scheme based on the bipartite graph matching model, this application further includes: adaptive TDMA time slot allocation for target user nodes that have been allocated independent channels; adaptive TDMA time slot allocation for target user nodes that have been allocated independent channels includes: S1, dividing a preset time domain into multiple transmission time slots of equal length; S2, multiple target user nodes sending their own data queue length and remaining energy value to the aggregation gateway; S3, the aggregation gateway calculating the time slot allocation weight based on the data queue length and remaining energy value, and allocating a transmission time slot within the preset time domain to each target user node according to the time slot allocation weight (so that each target user node can use the allocated independent channel to send data to the aggregation gateway within the allocated transmission time slot).

[0050] Before calculating the time slot allocation weight, this application further includes: the aggregation gateway determining whether the remaining energy value of each target user node is lower than a preset energy warning threshold; if the remaining energy value is lower than the energy warning threshold (indicating that the user node has insufficient energy), the time slot allocation weight of the corresponding target user node is set to 0, and it is determined that no transmission time slots will be allocated to the target user node in at least one time domain after the preset time domain (i.e., allocation is temporarily suspended to avoid device failure due to excessively low energy); if it is not lower than the preset energy warning threshold, the weight is calculated normally and time slots are allocated. This operation before calculating the time slot allocation weight can be understood as the time slot allocation adjustment after the energy warning. If the preset energy warning threshold is 65%, the remaining energy of each user node is determined as follows: Node1 (80%≥65%), Node2 (75%≥65%), Node3 (70%≥65%), Node4 (65%=65%), and Node5 (60%<65%). Therefore, the time slot allocation weight of Node5 is set to 0, and no transmission time slots are allocated to Node5 within the two time domains (a total of 200ms) after the current preset time domain (100ms), allowing it to enter the energy-saving mode and ensuring the device's battery life.

[0051] The aforementioned TDMA time slot allocation is a Time Division Multiple Access (TDMA) time slot allocation. It divides the preset time domain into multiple equal-length transmission time slots, allocating a dedicated time slot to each user node on an assigned channel. This enables time-division multiplexing of multiple user nodes on the same channel, avoiding time slot conflicts. It further optimizes time domain resources based on the allocation of independent channels to user nodes, preventing signal conflicts caused by simultaneous transmission from different user nodes on the same channel, improving channel time utilization, and dynamically adjusting time slots according to the data requirements and energy consumption status of user nodes to achieve a balance between energy saving and efficient transmission. For example, the preset time domain is divided into 10 equal-length transmission time slots, each with a duration of 10ms, for a total time domain duration of 100ms. Based on the data queue length and remaining energy of the user nodes, time slots 1-2 are allocated to Node1, time slots 3-4 to Node2, and so on, ensuring that each user node transmits data within its dedicated time slot. The longer the data queue length, the greater the weight of the time slot allocation; the lower the remaining energy value, the smaller the weight of the time slot allocation.

[0052] S1 above, which involves dividing the transmission time slots, can be understood as follows: A preset time domain is divided into multiple transmission time slots of equal length. The time slot length is set according to the data transmission cycle and latency requirements of the user nodes. The time slot length must ensure that a single user node completes one data transmission within that time slot. For example, if the preset time domain is 100ms, it is divided into 10 transmission time slots of equal length, each time slot lasting 10ms, numbered 1 to 10. Only one user node is allowed to transmit data in each time slot to avoid time slot conflicts. S2 above refers to user node feedback data. Specifically, target user nodes allocated independent channels send their core parameters to the aggregation gateway: data queue length (the amount of data currently to be transmitted) and remaining energy value (the remaining battery power of the device). These two parameters are the core basis for the aggregation gateway to calculate the time slot allocation weight. The data returned by each target user node is as follows: Node1 (data queue length 800 bits, remaining energy 80%), Node2 (data queue length 600 bits, remaining energy 75%), Node3 (data queue length 500 bits, remaining energy 70%), Node4 (data queue length 400 bits, remaining energy 65%), Node5 (data queue length 300 bits, remaining energy 60%). S3 calculates the weights for time slot allocation and allocates time slots. Specifically, the aggregation gateway calculates the time slot allocation weight for each user node based on its data queue length and remaining energy value using a preset algorithm (the larger the data queue length and the higher the remaining energy, the greater the weight and the more time slots allocated). Based on the weight ratio, a transmission time slot within a preset time domain is allocated to each user node. For example, the formula for calculating the time slot allocation weight is: Weight = (Data queue length / Total data queue length) × 0.6 + (Remaining energy / Total remaining energy) × 0.4; Total data queue length = 800 + 600 + 500 + 400 + 300 = 2600 bits, Total remaining energy = 80% + 75% + 70% + 65% + 60% = 350%; Calculate the weights of each node: Node1≈0.24 + 0.091≈0.331, Node2≈0.18 + 0.086≈0.266, Node3≈0.15 + 0.08≈0.23, Node4≈0.12 + 0.074≈0.194, Node5≈0.09 + 0.069≈0.159; Allocate 10 time slots according to the weight ratio: Node1 is allocated 3 time slots (1~3), Node2 is allocated 3 time slots (4~6), Node3 is allocated 2 time slots (7~8), Node4 is allocated 1 time slot (9), and Node5 is allocated 1 time slot (10).

[0053] Underwater acoustic channels are characterized by narrow available bandwidth, high propagation delay, and high collision rate. The function of the MAC layer protocol is to solve the channel allocation and access problems of nodes, thereby reducing collisions and improving network throughput. In clustered networks, multiple cluster member nodes (CMNs) often need to send data packets to a single cluster head node (CHN). To avoid data packet collisions, the TDMA protocol can be used. TDMA is a fixed time slot allocation protocol that divides time into multiple time frames, each frame with a fixed number of time slots, and then allocates these time slots to each node. Nodes send data according to their allocated time slots. However, if a node does not need to send data, its time slots are wasted. If some nodes need to send a large amount of data, but the allocated number of time slots is insufficient, the node's sending demand cannot be met, excessive data is buffered, resulting in excessive queue delays, and even data overflow leading to the loss of important data. Conversely, if some nodes have too little remaining energy, they should conserve energy to send higher-priority data. In this case, the number of time slots allocated to them should also be reduced appropriately; otherwise, it will cause continuous energy consumption, thus affecting the overall network performance.

[0054] The real-time method described in this application solves the problem of low channel utilization caused by fixed time slot allocation in TDMA. This application adopts an improved TDMA algorithm, which can perform efficient and adaptive time slot allocation, thus solving the problem of low channel utilization. At the same time, it improves network throughput and reduces network packet loss rate.

[0055] After determining the target allocation scheme, this application further includes: S1, the target user node that has been allocated an independent channel periodically measures the Doppler frequency shift of the received signal and feeds it back to the aggregation gateway; S2, when the Doppler frequency shift of any target user node exceeds a preset Doppler tolerance threshold, the aggregation gateway recalculates the expected channel capacity of multiple user nodes and determines an updated matching scheme. The measurement of the Doppler frequency shift of the received signal in S1 includes: S11, the aggregation gateway sends a probe acoustic wave signal of a preset frequency to each target user node; S12, each target user node receives the probe acoustic wave signal and determines the difference between the received frequency and the transmitted frequency as the Doppler frequency shift.

[0056] The aforementioned Doppler frequency shift refers to the difference between the frequency of the received signal and the frequency of the transmitted signal caused by the relative movement between the user node and the aggregation gateway (such as the movement of the user node due to water flow or minor adjustments in the position of the aggregation gateway). It is an important indicator for measuring the stability of underwater communication links. It is used to monitor dynamic changes in the communication link. When the Doppler frequency shift exceeds a preset threshold, it indicates a decrease in link stability, requiring a recalculation of the expected channel capacity and an update of the channel allocation scheme to ensure that communication quality is not affected. S1 to S2 can be understood as the channel allocation trigger monitoring process. When the Doppler frequency shift of any target user node exceeds the preset Doppler tolerance threshold, the aggregation gateway determines that the position of the target user node has changed significantly, and then re-allocates the channel to obtain an updated matching scheme. For example, in S1 to S2 above: the preset measurement period is 10s, each target user node measures the Doppler frequency shift of the received signal every 10s and feeds it back to GW1; the feedback results of a certain measurement are as follows: Node1 (0.5Hz), Node2 (0.8Hz), Node3 (1.0Hz), Node4 (1.2Hz), Node5 (0.6Hz).

[0057] Furthermore, when the channel establishment and allocation method for unmanned platforms in this application is adopted, the aggregation gateway is deployed as the cluster network, the aggregation gateway acts as the cluster head node, and multiple user nodes act as cluster member nodes. Each cluster member node communicates directly only with the cluster head node of its own cluster. Different clusters use different frequency sub-bands within the 10kHz to 40kHz spectrum to avoid mutual interference.

[0058] The channel establishment and allocation method for unmanned platforms described in this application achieves efficient channel allocation between five user nodes and the aggregation gateway. The target matching scheme ensures that the minimum expected channel capacity (0.5 kbps) for all user nodes is maximized, meeting the basic data transmission requirements of each node. Adaptive TDMA time slot allocation dynamically adjusts based on data volume and energy consumption, improving time domain resource utilization, while protecting low-power nodes through energy warning. Doppler frequency shift monitoring enables dynamic adaptation, ensuring communication stability in dynamic underwater environments. Clustered deployment avoids inter-cluster interference, improving the communication efficiency of the entire underwater network and fully meeting the communication requirements for shallow marine environmental monitoring. This application utilizes user location knowledge to maximize the minimum channel capacity achieved by each user. This provides maximum fairness and makes more efficient use of available spectrum resources. Performance evaluation through simulation shows that, compared to a fixed allocation scheme, the dynamic allocation method can significantly improve fairness among users and can be better scaled, thereby achieving effective communication over greater distances.

[0059] According to another aspect of the present invention, a channel establishment and allocation apparatus for unmanned platforms is also provided for implementing the above-described channel establishment and allocation method for unmanned platforms, such as... Figure 3 As shown, the device includes: an acquisition unit 302, used to acquire a core available frequency band, multiple user nodes, a convergence gateway, the number of nodes of the multiple user nodes, and the communication distance from each user node to the convergence gateway, wherein the multiple user nodes and the convergence gateway are unmanned platform nodes; a calculation unit 304, used to divide the core available frequency band into multiple independent channels of equal width according to the number of nodes, and to calculate the expected channel capacity of each user node on each independent channel based on the multiple communication distances using a communication attenuation model; a construction unit 306, used to construct a bipartite graph matching model based on the multiple user nodes, multiple independent channels, and multiple expected channel capacities, wherein the first vertex set of the bipartite graph matching model is the user node set, the second vertex set is the independent channel set, and the edge weight between the first vertex and the second vertex is the expected channel capacity of the corresponding user node on the corresponding independent channel; and a determination unit 308, used to determine a target matching scheme based on the bipartite graph matching model, wherein the target matching is the matching scheme that maximizes the minimum expected channel capacity of the multiple user nodes, and to allocate a unique independent channel to each user node according to the target matching scheme.

[0060] The specific methods of execution of each unit in the above device embodiments have been described in detail in the embodiments related to the method, and will not be elaborated further here.

[0061] According to another aspect of the present invention, an electronic device for implementing the above-described channel establishment and allocation method for unmanned platforms is also provided. This electronic device may be a terminal device or a server. This embodiment uses a terminal device as an example. The electronic device may include: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor to cause the at least one processor to perform the steps in any of the above-described embodiments of the channel establishment and allocation method for unmanned platforms.

[0062] According to one aspect of this application, a computer-readable storage medium is provided, from which a processor of a computer device reads computer instructions and executes the computer instructions, causing the computer device to perform the aforementioned channel establishment and allocation method for unmanned platforms.

[0063] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A channel establishment and allocation method for unmanned platforms, characterized in that, include: The system obtains the core available frequency bands, multiple user nodes, a convergence gateway, the number of nodes of the multiple user nodes, and the communication distance from each user node to the convergence gateway. The multiple user nodes and the convergence gateway are unmanned platform nodes. Based on the number of nodes, the core available frequency band is divided into multiple independent channels of equal width, and a communication attenuation model is used to calculate the expected channel capacity of each user node on each independent channel according to the multiple communication distances. A bipartite graph matching model is constructed based on multiple user nodes, multiple independent channels, and multiple expected channel capacities. The first vertex set of the bipartite graph matching model is the user node set, the second vertex set is the independent channel set, and the edge weight between the first vertex and the second vertex is the expected channel capacity of the corresponding user node on the corresponding independent channel. The target matching scheme is determined according to the bipartite graph matching model. The target matching is a matching scheme that maximizes the minimum expected channel capacity of multiple user nodes, and a unique independent channel is allocated to each user node according to the target matching scheme.

2. The method according to claim 1, characterized in that, Using a communication attenuation model, the expected channel capacity of each user node on each independent channel is calculated based on multiple communication distances, including: The communication attenuation model is established based on the pitch attenuation model and the Thorp formula, wherein the Thorp formula is used to calculate the seawater absorption loss of underwater acoustic propagation. Using a tone attenuation model, the total propagation loss after monotonic audio propagation for each of the multiple communication distances and the seawater absorption loss is calculated. The noise power spectrum is calculated based on the obtained turbulence noise components, shipping noise components, wind and wave noise components, and thermal noise components. The expected channel capacity is calculated using the channel capacity theorem based on the noise power spectrum and the total propagation loss.

3. The method according to claim 1, characterized in that, Determining a target matching scheme based on the bipartite graph matching model includes performing an iterative deletion operation based on the bipartite graph matching model, and determining the target matching scheme based on the iterative deletion result. The iterative deletion operation includes: Multiple vertex edges between the first vertex set and the second vertex set corresponding to the bipartite graph matching model are determined as a candidate edge set, and the candidate edge corresponding to the smallest edge weight is determined as the smallest weight edge according to the edge weight corresponding to each candidate edge. Remove the minimum weight edge from the candidate edge set to obtain a reference candidate edge set, and determine whether the reference candidate edge set satisfies the reference matching requirement, which is used to indicate the matching that can allocate different independent channels to each user node; If the reference matching requirement is met, the reference candidate edge set is determined as the candidate edge set and the iteration deletion operation continues. If the reference matching requirement is not met, the minimum weight edge is determined as the locked allocation edge, and the weight value of the locked allocation edge is determined as the allocation capacity threshold. Remove candidate edges from the candidate edge set that correspond to the same user node or the same channel as the locked allocation edge. When the number of locked allocation edges is equal to the number of user nodes, the iterative deletion operation terminates, thereby determining the target matching scheme.

4. The method according to claim 3, characterized in that, After determining the locked allocation edge, the method further includes: removing candidate edges from the candidate edge set whose edge weights are lower than the weight value corresponding to the locked allocation edge.

5. The method according to claim 1, characterized in that, After determining the target allocation scheme based on the bipartite graph matching model, the method further includes: adaptive TDMA time slot allocation for target user nodes that have been allocated independent channels; The adaptive TDMA time slot allocation for target user nodes that have been allocated independent channels includes: dividing the preset time domain into multiple transmission time slots of equal length; Multiple target user nodes send their own data queue length and remaining energy value to the aggregation gateway; The aggregation gateway calculates the time slot allocation weight based on the data queue length and the remaining energy value, and allocates the transmission time slot within the preset time domain to each target user node according to the time slot allocation weight.

6. The method according to claim 5, characterized in that, Before calculating the time slot allocation weights, the following is also included: The aggregation gateway determines whether the remaining energy value of each target user node is lower than a preset energy warning threshold; If the remaining energy value is lower than the energy warning threshold, the time slot allocation weight corresponding to the target user node is set to 0, and it is determined that no transmission time slot will be allocated to the target user node in at least one time domain after the preset time domain.

7. The method according to claim 1, characterized in that, After determining the target allocation plan, the following is also included: The target user node, which has been assigned an independent channel, periodically measures the Doppler frequency shift of the received signal and feeds it back to the aggregation gateway; When the Doppler frequency shift of any target user node exceeds a preset Doppler tolerance threshold, the aggregation gateway recalculates the expected channel capacity of the multiple user nodes and determines an updated matching scheme.

8. The method according to claim 7, characterized in that, Measuring the Doppler frequency shift of the received signal includes: The aggregation gateway sends a detection acoustic wave signal of a preset frequency to each of the target user nodes; Each target user node receives the probe acoustic signal and determines the difference between the received frequency and the transmitted frequency as the Doppler frequency shift.

9. The method according to claim 1, characterized in that, When the aggregation gateway is deployed as the clustered network, the aggregation gateway acts as the cluster head node, and the multiple user nodes act as cluster member nodes. Each cluster member node communicates directly only with the cluster head node of its own cluster, and different clusters use different frequency sub-bands within the 10kHz to 40kHz spectrum.

10. A channel establishment and allocation device for unmanned platforms, characterized in that, include: The acquisition unit is used to acquire the core available frequency band, multiple user nodes, aggregation gateway, the number of nodes of the multiple user nodes, and the communication distance from each user node to the aggregation gateway, wherein the multiple user nodes and the aggregation gateway are unmanned platform nodes; The computing unit is used to divide the core available frequency band into multiple independent channels of equal width according to the number of nodes, and to calculate the expected channel capacity of each user node on each independent channel based on the multiple communication distances using a communication attenuation model. The construction unit is used to construct a bipartite graph matching model based on multiple user nodes, multiple independent channels, and multiple expected channel capacities. The first vertex set of the bipartite graph matching model is the user node set, the second vertex set is the independent channel set, and the edge weight between the first vertex and the second vertex is the expected channel capacity of the corresponding user node on the corresponding independent channel. The determining unit is configured to determine a target matching scheme based on the bipartite graph matching model, wherein the target matching is a matching scheme that maximizes the minimum expected channel capacity of multiple user nodes, and to allocate a unique independent channel to each user node according to the target matching scheme.

Images