Unmanned aerial vehicle communication protocol feature extraction method and system based on artificial intelligence, electronic equipment and storage medium

Abstract

The invention provides an unmanned aerial vehicle communication protocol feature extraction method and system based on artificial intelligence, electronic equipment and a storage medium, and relates to the technical field of unmanned aerial vehicle communication.The method comprises the steps that a communication environment model is constructed by obtaining a spatial position set of a base station and an unmanned aerial vehicle group and synchronously collecting real-time speed and acceleration information of all position points; mapping to generate a virtual space shape, calculating an offset, performing motion compensation to generate a correction parameter, and performing closed-loop adjustment on a shape parameter; then, deep correlation features are extracted, screening parameters are generated, key features are screened, and protocol layer compensation coefficients are generated through fusion with signal attenuation model parameters; and finally, extracting spatio-temporal change features based on the position relation features, predicting a communication interruption path, extracting protocol features in combination with related parameters, evaluating risks, and generating a standby node dynamic selection strategy, so that unmanned aerial vehicle communication protocol feature extraction and standby communication node dynamic selection based on artificial intelligence can be realized.

Description

Technical Field

[0001] The present application relates to the field of drone communication technology, and in particular to an artificial intelligence-based drone communication protocol feature extraction method, system, electronic device, and storage medium. Background Art

[0002] In complex scenarios where high-speed movement triggers sudden changes in three-dimensional topology, there is an urgent need for drone-based self-organizing network relay communications. In emergency rescue scenarios, drones must navigate complex terrain, such as mountains and ruins, and rapidly adjust communication links. This requires communication protocols that can perceive and respond to topological changes in real time. Furthermore, with a large number of drones operating simultaneously, limited resources such as bandwidth must be allocated rationally to meet the diverse quality of service requirements for different tasks in terms of data transmission rate and latency.

[0003] One currently available solution is to use a communication protocol based on software-defined networking (SDN). This protocol leverages a centralized controller to globally control network topology and flexibly allocate network resources. The controller collects information such as the location and link quality of each drone node and dynamically adjusts routing strategies based on the real-time network status. This allows for efficient communication link construction and management, adapting to the rapid topology changes caused by the high-speed movement of drone nodes.

[0004] However, this solution has significant flaws. A failure in the centralized controller risks paralyzing the entire communication network, resulting in poor reliability and stability. Furthermore, as the number of drones and network size increase, the controller must process massive amounts of information, placing a significant strain on computing and storage. This can lead to processing delays and an inability to respond promptly to sudden changes in three-dimensional topology, resulting in decreased communication quality. This makes it difficult to meet the stringent requirements for real-time and reliable communication in complex scenarios. Summary of the Invention

[0005] The purpose of this application is to provide an artificial intelligence-based drone communication protocol feature extraction method, system, electronic device and storage medium to solve the problem in the existing technology that it is difficult to extract communication protocol features in real time and accurately and dynamically optimize the selection of backup nodes in scenarios where the high-speed movement of drones causes three-dimensional topological mutations.

[0006] To solve the above technical problems, in the first aspect, the present application provides a method for extracting UAV communication protocol features based on artificial intelligence, comprising: Acquire a spatial position set of the base station and the drone swarm, synchronously collect real-time velocity and acceleration information of each position point in the spatial position set, and construct a communication environment model based on the spatial position set, real-time velocity and acceleration information; Mapping the set of spatial positions to generate a virtual spatial shape, calculating a spatial offset between a geometric center point of the virtual spatial shape and a center point of a communication range, performing motion compensation on the spatial offset in combination with the acceleration information to generate a protocol feature correction parameter, and adjusting shape parameters of the virtual spatial shape through feedback of the protocol feature correction parameter to form a closed-loop geometric modeling; Extracting deep correlation features of the communication environment model, generating feature screening parameters based on the dimensional distribution characteristics of the communication environment model, screening key features of the deep correlation features based on the feature screening parameters, and fusing the screened key features with preset signal attenuation model parameters to generate protocol layer compensation coefficients; Based on the positional relationship features output by the closed-loop geometric modeling, the spatiotemporal variation features of the communication network are extracted, and the communication interruption path is predicted based on the spatiotemporal variation features. Based on the protocol feature correction parameters, the protocol layer compensation coefficient and the communication interruption path, the communication protocol features are extracted and the communication interruption risk probability is evaluated to generate a dynamic selection strategy for backup communication nodes at the communication protocol layer.

[0007] Optionally, extracting the communication protocol features and evaluating the communication interruption risk probability based on the protocol feature correction parameter, the protocol layer compensation coefficient, and the communication interruption path to generate a dynamic selection strategy for a backup communication node at the communication protocol layer includes: The module length of the protocol feature correction parameter, the protocol layer compensation coefficient, the azimuth angle and the pitch angle of the communication interruption path are combined into a communication protocol feature vector, and the communication protocol feature vector is input into the pre-trained risk probability calculation function to output the communication interruption risk probability of each drone node relative to the base station; Sort the communication interruption risk probabilities of all drone nodes in ascending order, and select the top N nodes with the lowest communication interruption risk probability to form the backup communication node set; Continuously monitor the quality parameters of the primary communication link. When the quality parameters are lower than a first quality threshold, switch the communication routing node to the node with the highest priority in the backup node set. Detect the quality parameters of the new communication link within a fixed verification time window after the switch. If the quality parameters of the new communication link do not recover to a second quality threshold, remove the current node from the backup node set and switch back to the node with the highest priority among the remaining nodes. When there is no available node in the standby node set, the interruption risk probability calculation is re-executed to generate a new standby node set.

[0008] Optionally, adjusting shape parameters of the virtual space shape by feedback of the protocol feature correction parameters to form closed-loop geometric modeling includes: Determining a vertex movement direction and a movement distance of a virtual space shape based on the protocol feature correction parameter; performing a synchronous translation operation on all vertex coordinates of the virtual space shape according to the vertex movement direction and movement distance, and reconnecting the translated vertex coordinates to generate a new virtual space shape; Calculating the coordinates of the geometric center point of the new virtual space shape, and measuring the difference between the coordinates of the geometric center point and the coordinates of the center point of the base station communication range as a closed-loop verification offset; When the closed-loop verification offset is less than a preset threshold, the current virtual space shape is output as the final geometric model. When the closed-loop verification offset is greater than or equal to the preset threshold, the closed-loop verification offset is input as a new space offset into the motion compensation process to regenerate the protocol feature correction parameters.

[0009] Optionally, extracting spatiotemporal variation characteristics of the communication network based on the positional relationship characteristics output by the closed-loop geometric modeling, and predicting the communication interruption path according to the spatiotemporal variation characteristics, includes: Obtain the coordinates of all boundary vertices in the final geometric model output by closed-loop geometric modeling, and calculate the distances between adjacent vertices to form a positional relationship feature sequence; Analyze the position relationship feature sequence, extract the distance fluctuation range in each time window as the amplitude feature, and calculate the number of distance zero-crossing changes per unit time as the frequency feature, and combine the amplitude feature and the frequency feature to form the spatiotemporal change feature; In the spatiotemporal variation characteristics, boundary line segments that satisfy the requirement of continuously increasing amplitude and a frequency exceeding a preset frequency threshold are detected, and the extension direction vectors of these line segments are determined as candidate directions of the communication interruption path. Cluster analysis is performed on the candidate directions of the communication interruption path, and the direction of the maximum cluster center is taken as the communication interruption path.

[0010] Optionally, mapping the set of spatial positions to generate a virtual spatial shape, calculating a spatial offset between a geometric center point of the virtual spatial shape and a center point of a communication range, and performing motion compensation on the spatial offset in combination with the acceleration information to generate a protocol feature correction parameter includes: Inputting the position coordinates of the drones in the spatial position set into the spatial mapping rule, and forming a virtual spatial shape by connecting the position coordinates of the outermost drones; Calculating the arithmetic mean of all vertex coordinates of the virtual space shape as the coordinates of the geometric center point, and simultaneously measuring the spatial offset between the coordinates of the geometric center point and the coordinates of the center point of the base station communication range; Multiplying the acceleration information of each drone node by a preset time interval to obtain an instantaneous velocity change, multiplying the instantaneous velocity change by the same time interval to obtain a displacement compensation component, and performing vector superposition on the displacement compensation components of all drone nodes to generate an overall displacement compensation vector; A vector subtraction operation is performed on the spatial offset and the overall displacement compensation vector to obtain a protocol feature correction parameter.

[0011] Optionally, extracting the deep correlation features of the communication environment model, generating feature screening parameters according to the dimensional distribution characteristics of the communication environment model, screening key features of the deep correlation features based on the feature screening parameters, and fusing the screened key features with preset signal attenuation model parameters to generate protocol layer compensation coefficients includes: The distance change rate and speed correlation strength between all drone node pairs are extracted from the spatial topology network of the communication environment model as deep correlation features; Calculate the feature dispersion of the deep correlation feature in the three-dimensional coordinate direction, and take the maximum dispersion value in the three directions as the feature screening parameter; Screening out deep correlation features whose distance change rate is greater than the feature screening parameter and whose velocity correlation strength is greater than the feature screening parameter as key features; Multiplying the distance change rate of the key feature by a preset distance attenuation coefficient to obtain a distance compensation component; Multiplying the velocity correlation strength of the key feature by a preset motion interference coefficient to obtain a motion compensation component; The distance compensation component and the motion compensation component are weightedly superimposed to generate a protocol layer compensation coefficient.

[0012] Optionally, acquiring a spatial position set of a base station and a swarm of drones, synchronously collecting real-time velocity and acceleration information of each position point in the spatial position set, and constructing a communication environment model based on the spatial position set, real-time velocity and acceleration information, includes: Obtain the spatial position set of the base station and all individual drones in the drone group, and synchronously record the real-time speed and acceleration information corresponding to each position point in the spatial position set; Establishing a spatial topology network including base station nodes and drone nodes based on the spatial position set, and marking the real-time speeds of the base station nodes and drone nodes at corresponding positions in the spatial topology network; Building a dynamic connection relationship in the spatial topology network according to the change in spatial position distance and real-time speed difference between adjacent nodes; The acceleration information is integrated into the dynamic connection relationship as a motion state variable to form a communication environment model that can be updated in real time.

[0013] In a second aspect, the present application provides an artificial intelligence-based drone communication protocol feature extraction system, comprising: An acquisition module is configured to acquire a spatial position set of the base station and the drone swarm, synchronously collect real-time velocity and acceleration information of each position point in the spatial position set, and construct a communication environment model based on the spatial position set, real-time velocity and acceleration information; a compensation module, configured to map the set of spatial positions to generate a virtual spatial shape, calculate a spatial offset between a geometric center point of the virtual spatial shape and a center point of a communication range, perform motion compensation on the spatial offset in combination with the acceleration information to generate protocol feature correction parameters, and adjust shape parameters of the virtual spatial shape using feedback from the protocol feature correction parameters to form a closed-loop geometric modeling; a screening module, configured to extract deep correlation features of the communication environment model, generate feature screening parameters based on the dimensional distribution characteristics of the communication environment model, screen key features of the deep correlation features based on the feature screening parameters, and fuse the screened key features with preset signal attenuation model parameters to generate protocol layer compensation coefficients; A generation module is used to extract the spatiotemporal variation characteristics of the communication network based on the positional relationship characteristics output by the closed-loop geometric modeling, predict the communication interruption path according to the spatiotemporal variation characteristics, extract the communication protocol characteristics based on the protocol characteristic correction parameters, protocol layer compensation coefficients and communication interruption path, and evaluate the communication interruption risk probability, so as to generate a dynamic selection strategy for backup communication nodes in the communication protocol layer.

[0014] In a third aspect, the present application provides an electronic device, comprising: memory for storing computer programs; A processor is configured to implement the steps of the method for extracting features of drone communication protocols based on artificial intelligence as described in the first aspect above when executing the computer program.

[0015] In a fourth aspect, the present application provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, it can implement the steps of the artificial intelligence-based drone communication protocol feature extraction method as described in the first aspect above.

[0016] The artificial intelligence-based drone communication protocol feature extraction method provided in this application constructs a communication environment model by acquiring the spatial position and motion information of the base station and the drone group, which can accurately depict the communication scene; realizes dynamic adjustment of the virtual space shape through closed-loop geometric modeling, and improves the compensation ability for position offset caused by high-speed movement of the drone; generates protocol layer compensation coefficients by extracting deep correlation features and fusing signal attenuation model parameters, thereby enhancing the pertinence and accuracy of feature extraction; extracts spatiotemporal change features based on position relationship features and predicts the communication interruption path, and combines relevant parameters to extract protocol features and evaluate risks, which can effectively respond to three-dimensional topological mutations. The dynamic selection strategy for backup communication nodes finally generated can significantly improve the real-time, reliability and anti-interference capability of drone communications, and ensure the stability and efficiency of communication links in complex dynamic scenarios.

[0017] Furthermore, the module length of the protocol feature correction parameters, the protocol layer compensation coefficient, and the azimuth and pitch angles of the communication interruption path are combined into a communication protocol feature vector. This is then input into a pre-training function to obtain the communication interruption risk probability of each drone node. The probability is then sorted in ascending order, and the top N nodes are selected to form a backup node set. The quality of the main link is then monitored. If it falls below a first threshold, the node with the highest priority in the backup set is switched to. If the new link quality does not reach a second threshold within a fixed time, the node is removed and switched to the next one. If no nodes are available, a new set is recalculated. This method accurately assesses node interruption risk by constructing a feature vector, dynamically screens backup nodes, and combines this with main link quality monitoring to achieve intelligent node switching and updating. This method can quickly respond to changes in link quality, effectively ensure communication continuity, and improve the stability and anti-interruption capabilities of drone communication networks in dynamic scenarios. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the embodiments of the present application or the technical solutions of the prior art, the following is a brief introduction to the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0019] Figure 1 A flowchart of a method for extracting features from a drone communication protocol based on artificial intelligence provided by an embodiment of the present application; Figure 2 A flowchart of a specific implementation of a method for extracting features of drone communication protocols based on artificial intelligence provided in an embodiment of the present application; Figure 3 A diagram showing a specific implementation scenario of a method for extracting features of drone communication protocols based on artificial intelligence provided in an embodiment of the present application; Figure 4A schematic diagram of the structure of an artificial intelligence-based drone communication protocol feature extraction system provided in an embodiment of the present application. DETAILED DESCRIPTION

[0020] Existing communication protocols based on software-defined networking (SDN) have significant shortcomings in the three-dimensional topology change scenarios of drone ad hoc relay communications. A failure of the centralized controller can paralyze the network, resulting in poor reliability. Furthermore, as the number of drones and network scale increase, controllers are prone to delays in processing massive amounts of information, making it difficult to respond to topology changes in a timely manner and meeting the real-time and reliability requirements of communication. In response to the above problems, this application proposes a method for extracting UAV communication protocol features based on artificial intelligence. This method constructs a communication environment model by acquiring the spatial position, velocity, and acceleration information of the base station and the UAV group, dynamically adjusts the shape of the virtual space through closed-loop geometric modeling, extracts deep correlation features, and fuses the signal attenuation model parameters to generate compensation coefficients. The position relationship features are then combined to extract spatiotemporal variation features and predict communication interruption paths. Finally, protocol features are extracted and risks are assessed to generate a dynamic selection strategy for backup nodes. This solution does not rely on a centralized controller, can accurately extract protocol features in real time, and dynamically optimize backup node selection. It effectively solves the problems of poor reliability, processing delays, and difficulty in coping with topology mutations in existing solutions, and improves the real-time and stability of communications.

[0021] In order to enable those skilled in the art to better understand the present application, the present application is further described in detail below in conjunction with the accompanying drawings and specific embodiments. Obviously, the embodiments described are only a part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without making any creative efforts are within the scope of protection of the present application.

[0022] The core of this application is to provide a method for extracting features of drone communication protocols based on artificial intelligence. The flowchart of a specific implementation method is as follows: Figure 1 As shown, the method includes: S101, obtaining a spatial position set of a base station and a group of drones, synchronously collecting real-time velocity and acceleration information of each position point in the spatial position set, and constructing a communication environment model based on the spatial position set, real-time velocity and acceleration information; Optionally, step S101 may specifically include the following steps: S1011. Obtain a spatial position set of the base station and all individual drones in the drone group, and synchronously record real-time velocity and acceleration information corresponding to each position point in the spatial position set; S1012: Establish a spatial topology network including base station nodes and drone nodes based on the spatial position set, and mark the real-time speeds of the base station nodes and drone nodes at corresponding positions in the spatial topology network; S1013, constructing a dynamic connection relationship in the spatial topology network according to the spatial position distance change and the real-time speed difference between adjacent nodes; S1014: Integrate the acceleration information into the dynamic connection relationship as a motion state variable to form a communication environment model that can be updated in real time.

[0023] In the above scheme, the spatial location set refers to an information set that reflects the spatial distribution of the base station and each drone in the drone swarm. It contains the three-dimensional coordinate data of each location point and can be used to establish spatial associations between nodes. Real-time speed refers to dynamic information that reflects the current movement status of each location point. It contains the speed and direction parameters of movement and can be used to analyze the movement trend of the node. Acceleration information refers to a dynamic signal that reflects the speed change characteristics of each location point. It contains the speed and direction parameters of the speed change and can be used to predict the position change of the node. The communication environment model refers to a dynamic model constructed by integrating the spatial location set, real-time speed and acceleration information. It contains the connection relationship and movement state association between the base station and drone nodes and can be used to characterize the dynamic characteristics of the communication scenario.

[0024] In the embodiment of the present application, first, step S1011 uses the GPS module onboard the drone and the positioning system built into the base station to obtain real-time spatial position data for the base station and each drone in the drone group, forming a spatial position set. This data is presented in the form of three-dimensional coordinates (x, y, z). Simultaneously, using the speed sensors and accelerometers on the drone and base station, real-time speed and acceleration information corresponding to each location point is synchronously collected. Real-time speed includes speed magnitude and direction of movement, and acceleration also includes acceleration magnitude and direction of change. This information is temporarily stored in the data acquisition terminal to provide raw data for subsequent steps.

[0025] Next, in step S1012, based on the acquired spatial location set, a graph theory algorithm is used to treat the base station and each drone as network nodes. A spatial topology network is constructed in virtual space based on their three-dimensional coordinates, with the node locations in the network corresponding to their actual spatial locations. Subsequently, using data mapping technology, the real-time speed information of the base station and drone nodes collected in step S1011 is annotated next to the corresponding nodes in the spatial topology network according to the node correspondence. This allows the location and current speed of each node to be intuitively viewed in the topology network.

[0026] Next, in step S1013, based on the spatial topology network constructed in step S1012, the spatial position distance between two adjacent nodes at two consecutive time points is calculated using a time series analysis method to obtain the distance change. A positive value indicates an increase in distance, and a negative value indicates a decrease in distance. Simultaneously, the difference in magnitude and direction between the real-time speeds of the two adjacent nodes is calculated. The greater the speed difference, the more out of sync the two motion states are. Then, combining the distance change and speed difference, a dynamic connection algorithm is used to construct a dynamic connection relationship between nodes in the spatial topology network. When the distance change is small and the speed difference is small, the connection relationship is strong; otherwise, it is weak.

[0027] Finally, the dynamic connection relationship obtained in step S1013 is further processed in step S1014. Specifically, the acceleration information collected in step S1011 is used as a motion state variable and integrated into the dynamic connection relationship through data fusion technology. Because acceleration reflects the changing trend of speed, when the acceleration of a node changes, its future position and speed will change accordingly, which in turn affects the connection relationship between nodes. By introducing the acceleration factor into the dynamic connection relationship and applying a real-time update algorithm, the connection relationship can be automatically adjusted as the acceleration changes, ultimately forming a communication environment model that can be updated in real time. This model can reflect the dynamic changes in the spatial position, motion state, and node connection relationship between the base station and the drone group in real time.

[0028] In actual applications, in a drone inspection mission in a certain area, there is a base station P and a drone group consisting of drones Q1, Q2, and Q3. The three-dimensional coordinates of the base station P are obtained through the GPS module carried by the drone as (100, 200, 50), the coordinates of drone Q1 are (150, 220, 60), Q2 are (130, 180, 55), and Q3 are (170, 210, 65). At the same time, the speed sensor and acceleration sensor collect the speed of the base station P as 0 and the acceleration as 0, the speed of Q1 as (5, 3, 2) m / s and the acceleration as (1, 0.5, 0.3) m / s², the speed of Q2 as (4, -2, 1) m / s and the acceleration as (0.8, -0.4, 0.2) m / s², and the speed of Q3 as (6, 2, 3) m / s and the acceleration as (1.2, 0.6, 0.4) m / s². The data is stored in the data acquisition terminal. A graph theory algorithm is then used to construct a spatial topology network in virtual space based on the three-dimensional coordinates of base station P and Q1, Q2, and Q3. Data mapping technology is then used to annotate the speed information of each node next to the corresponding node. Based on this topological network, time series analysis is used to calculate that the change in distance between Q1 and Q2 at two consecutive time points is -3 meters, indicating a small speed difference. A dynamic connection algorithm is used to establish a strong connection between them. However, the change in distance between Q1 and Q3 is 5 meters, indicating a large speed difference and a weaker connection. Finally, data fusion technology is used to integrate the acceleration information of each node into the dynamic connection relationship. When Q2's acceleration changes to (0.7, -0.5, 0.1) m / s², the real-time update algorithm automatically adjusts its connection relationship with surrounding nodes, ultimately forming a communication environment model that can be updated in real time. The S101 solution, described above, constructs a communication environment model by acquiring the spatial locations of base stations and drone swarms and simultaneously collecting real-time velocity and acceleration information at each location. This model comprehensively and dynamically presents the spatial distribution, motion state, and inter-node connectivity of base station and drone nodes. By integrating and updating position, velocity, acceleration, and other information in real time, this model provides accurate and realistic foundational data for subsequent closed-loop geometric modeling and feature extraction steps. This makes the entire communication protocol feature extraction process more adaptable to dynamic conditions such as high-speed drone movement, effectively improving the perception of the communication environment and the accuracy of subsequent processing.

[0029] S102: Mapping the set of spatial positions to generate a virtual spatial shape, calculating a spatial offset between a geometric center point of the virtual spatial shape and a center point of a communication range, performing motion compensation on the spatial offset in combination with the acceleration information to generate protocol feature correction parameters, and adjusting shape parameters of the virtual spatial shape using feedback from the protocol feature correction parameters to form a closed-loop geometric modeling. Optionally, step S102 may specifically include the following steps: S1021: Input the position coordinates of the drones in the spatial position set into a spatial mapping rule, and form a virtual spatial shape by connecting the position coordinates of the outermost drones; S1022. Calculate the arithmetic mean of all vertex coordinates of the virtual space shape as the coordinates of the geometric center point, and simultaneously measure the spatial offset between the coordinates of the geometric center point and the coordinates of the center point of the base station communication range; S1023: Multiply the acceleration information of each UAV node by a preset time interval to obtain an instantaneous velocity change, multiply the instantaneous velocity change by the same time interval to obtain a displacement compensation component, and perform vector superposition on the displacement compensation components of all UAV nodes to generate an overall displacement compensation vector; S1024: Perform vector subtraction operation on the spatial offset and the overall displacement compensation vector to obtain a protocol feature correction parameter.

[0030] S1025, determining the vertex movement direction and movement distance of the virtual space shape based on the protocol feature correction parameter; S1026: performing a synchronous translation operation on all vertex coordinates of the virtual space shape according to the vertex movement direction and movement distance, and reconnecting the translated vertex coordinates to generate a new virtual space shape; S1027. Calculate the coordinates of the geometric center point of the new virtual space shape, and measure the difference between the coordinates of the geometric center point and the coordinates of the center point of the base station communication range as a closed-loop verification offset; S1028. When the closed-loop verification offset is less than a preset threshold, the current virtual space shape is output as the final geometric model; when the closed-loop verification offset is greater than or equal to the preset threshold, the closed-loop verification offset is input as a new space offset into the motion compensation process to regenerate the protocol feature correction parameters.

[0031] In the above scheme, the virtual spatial shape refers to a geometric figure generated by mapping the spatial positions of the drones, reflecting the distribution range of the drone swarm. It contains data such as vertex coordinates and is used to visually represent the spatial distribution of the drone swarm. The geometric center point is the point corresponding to the arithmetic mean of the coordinates of all vertices of the virtual spatial shape and represents the center position of the shape. The communication range center point is the center position of the base station's communication coverage area and is a fixed or known point. The spatial offset is the position difference between the geometric center point and the communication range center point. It contains three-dimensional distance information and is used to measure the positional deviation between the two. The instantaneous velocity change is the change in the speed of the drone within a preset time interval, reflecting the speed of the speed change. The displacement compensation component is calculated from the instantaneous velocity change and is used to compensate for the position offset. The overall displacement compensation vector is the vector formed by superimposing the displacement compensation components of all drones and is used to comprehensively reflect the impact of the drone swarm's motion on its position. The protocol feature correction parameter is a parameter obtained by calculating the spatial offset and the overall displacement compensation vector and is used to adjust the virtual spatial shape. Shape parameters, such as vertex coordinates, describe the shape of the virtual space. The closed-loop calibration offset is the difference between the geometric center of the new virtual space shape and the center of the communication range, used to verify the effectiveness of the adjustment. The preset threshold is the standard value used to determine whether the closed-loop calibration offset meets the requirements.

[0032] In the embodiment of the present application, the position coordinates of all drones in the spatial position set are first collected through step S1021, and a boundary detection algorithm is used to find the outermost drones. These drones are individuals at the edge of the drone group and are located relatively outside. Then, according to the spatial mapping rules, the position coordinates of the outermost drones are connected in sequence through geometric connection technology to form a closed virtual space shape. This shape can roughly outline the distribution range of the drone group. For example, if the position coordinates of the outermost drones are (2, 3, 4), (5, 6, 7), and (8, 9, 10), after connection, a triangular virtual space shape is formed.

[0033] Step S1022 is based on the virtual space shape obtained in step S1021. The coordinates of all vertices of the virtual space shape are first extracted. The arithmetic mean algorithm is used to calculate the average value of these vertex coordinates in the x, y, and z dimensions to obtain the coordinates of the geometric center point of the virtual space shape. At the same time, the coordinate difference calculation method is used to calculate the difference between the coordinates of the geometric center point and the coordinates of the center point of the pre-set base station communication coverage area in the x, y, and z directions. These differences are combined as the spatial offset. For example, if the coordinates of the geometric center point are (3, 4, 5) and the coordinates of the center point of the base station communication coverage area are (5, 5, 6), then the spatial offset is -2 in the x direction, -1 in the y direction, and -1 in the z direction.

[0034] Next, step S1023 uses the acceleration information of each drone node, first setting a preset time interval. Using kinematic formulas, the acceleration information of each drone is multiplied by the preset time interval to obtain the instantaneous velocity change of each drone during this time interval. This instantaneous velocity change is then multiplied by the same preset time interval to calculate the displacement compensation component for each drone. Then, using the vector superposition method, the displacement compensation components of all drones are accumulated in the x, y, and z directions. The resulting sum is the overall displacement compensation vector. For example, if a drone has an acceleration of (1, 2, 3) m / s² and a preset time interval of 2 seconds, its instantaneous velocity change is (2, 4, 6) m / s, and the displacement compensation components are (4, 8, 12) m. After superimposing the displacement compensation components of multiple drones, the overall displacement compensation vector is output.

[0035] Next, step S1024 processes the obtained spatial offset and overall displacement compensation vectors. Using vector subtraction, the overall displacement compensation vector is subtracted from the spatial offset vector. Specifically, the components of the two vectors in the x, y, and z directions are subtracted. The result is the protocol feature correction parameter, which reflects the magnitude and direction of the required adjustment to the virtual space shape. For example, if the spatial offset vector is (-2, -1, -1) and the overall displacement compensation vector is (10, 15, 20), then the protocol feature correction parameter is (-2 - 10, -1 - 15, -1 - 20) = (-12, -16, -21).

[0036] Then, in step S1025, the adjustment direction and movement distance are determined based on the obtained protocol feature correction parameters. The positive and negative values ​​of the protocol feature correction parameters in the x, y, and z directions respectively represent the movement direction of the virtual space shape vertex in the corresponding direction. A positive value indicates movement in the positive direction of that direction, while a negative value indicates movement in the negative direction. The magnitude of the component represents the distance the vertex needs to move. In step S1026, the coordinates of all vertices of the virtual space shape are processed according to the vertex movement direction and movement distance determined in step S1025. Using coordinate translation technology, the x, y, and z coordinates of each vertex are adjusted according to the corresponding movement direction and distance. Afterwards, all the translated vertex coordinates are reconnected in the original connection order to form a new virtual space shape. For example, if the vertex coordinates are (2, 3, 4), after adjusting according to the above movement direction and distance, they become (2-12, 3-16, 4-21) = (-10, -13, -17). The other vertices are similarly adjusted and reconnected.

[0037] Finally, in step S1027, the geometric center coordinates of all vertices of the new virtual space shape are calculated using an arithmetic mean algorithm. Then, using a coordinate difference calculation method, the differences between the new geometric center coordinates and the coordinates of the center point of the base station's communication range in the x, y, and z directions are calculated. These differences are combined to form the closed-loop calibration offset. Next, in step S1028, it is determined whether the obtained closed-loop calibration offset meets the requirements. A preset threshold is set, and the components of the closed-loop calibration offset in the x, y, and z directions are compared with the preset thresholds. If the absolute values ​​of all three components are less than the preset thresholds, the current virtual space shape meets the accuracy requirements and is output as the final geometric model. If the absolute value of any component is greater than or equal to the preset threshold, the closed-loop calibration offset is used as the new spatial offset and re-substituted into the motion compensation process initiated in step S1023. The instantaneous velocity change, displacement compensation component, and other parameters are recalculated to generate new protocol feature correction parameters. The virtual space shape is then adjusted again until the closed-loop calibration offset is less than the preset threshold.

[0038] In a practical application, in a certain drone collaborative operation scenario, there are base station P and drone groups, drones J, K, L, M, N. First, the position coordinates of these drones are collected. The boundary detection algorithm is used to determine that the outermost drones are J, L, and N. Their coordinates are (10, 10, 5), (20, 15, 5), and (15, 20, 5), respectively. According to the spatial mapping rules, these three coordinates are connected in sequence using geometric connection technology to form a virtual space shape. The coordinates of the vertices of the virtual space shape are extracted, and the coordinates of the geometric center point are calculated using the arithmetic mean algorithm. The coordinates of the center point of base station P's communication range are (18, 18, 5). Using the coordinate difference method, the spatial offsets are -3 in the x-direction, -3 in the y-direction, and 0 in the z-direction. Using the collected acceleration information from each drone, and setting a preset time interval of 1 second, drone J's acceleration is (2, 3, 0) m / s². Its instantaneous velocity change is (2, 3, 0) m / s², and the displacement compensation components are (2 × 1, 3 × 1, 0 × 1) = (2, 3, 0) m. Similarly, the displacement compensation components for K, L, M, and N are calculated. Using the vector superposition method, all displacement compensation components are accumulated to obtain the overall displacement compensation vector of (5, 6, 0) m. Performing vector subtraction between the spatial offset vector (-3, -3, 0) and the overall displacement compensation vector (5, 6, 0) yields the protocol feature correction parameters of (-3 - 5, -3 - 6, 0 - 0) = (-8, -9, 0). Then, based on the parameter, the vertex movement direction is determined to be the negative direction of the x direction and the negative direction of the y direction, and the movement distance is 8 units and 9 units respectively. According to this direction and distance, the coordinate translation technology is used to adjust the coordinates of the three vertices of the virtual space shape. The coordinates of J become (10-8, 10-9, 5-0) = (2, 1, 5), the coordinates of L become (20-8, 15-9, 5-0) = (12, 6, 5), and the coordinates of N become (15-8, 20-9, 5-0) = (7, 11, 5). The adjusted vertices are reconnected to form a new virtual space shape. Finally, the arithmetic mean algorithm is used to calculate the coordinates of the geometric center point of the new virtual space shape: , the closed-loop verification offset between the center point (18,18,5) of the base station communication range and the coordinate difference calculation method is -11 in the x-direction, -12 in the y-direction, and 0 in the z-direction; the preset threshold is set to 10. Since the absolute values ​​of the components of the closed-loop verification offset in the x- and y-directions are both greater than 10, the offset is substituted into the motion compensation process as the new spatial offset. After multiple adjustments, the final closed-loop verification offset is less than the preset threshold, and the virtual space shape at this time is output as the final geometric model.

[0039] The overall solution in S102 described above generates a virtual space shape and calculates spatial offsets. It then uses acceleration information to perform motion compensation to obtain protocol feature correction parameters. These parameters are then used as feedback to adjust the virtual space shape, forming a closed-loop geometric model. This process dynamically corrects deviations between the virtual space shape and the actual communication range caused by drone motion, making the constructed geometric model more consistent with the real-time spatial distribution of the drone swarm. This provides an accurate spatial geometry foundation for subsequent extraction of communication protocol features and enhances adaptability to high-speed drone movement scenarios.

[0040] S103: Extracting deep correlation features of the communication environment model, generating feature screening parameters based on dimensional distribution characteristics of the communication environment model, screening key features of the deep correlation features based on the feature screening parameters, and fusing the screened key features with preset signal attenuation model parameters to generate protocol layer compensation coefficients; Optionally, step S103 may specifically include the following steps: S1031. Extract the distance change rate and speed correlation strength between all drone node pairs from the spatial topology network of the communication environment model as deep correlation features; S1032, calculating the feature dispersion of the deep correlation feature in the three-dimensional coordinate direction, and taking the maximum dispersion value in the three directions as the feature screening parameter; S1033, screening out deep correlation features whose distance change rate is greater than the feature screening parameter and whose speed correlation strength is greater than the feature screening parameter as key features; S1034. Multiply the distance change rate of the key feature by a preset distance attenuation coefficient to obtain a distance compensation component; S1035: multiplying the velocity correlation strength of the key feature by a preset motion interference coefficient to obtain a motion compensation component; S1036: Perform weighted superposition on the distance compensation component and the motion compensation component to generate a protocol layer compensation coefficient.

[0041] In the above scheme, deep correlation features are extracted from the communication environment model and reflect the relationships between drone nodes. These include the distance change rate and velocity correlation strength between pairs of nodes, demonstrating the dynamic connection between nodes. The distance change rate refers to the rate of change in the distance between two drone nodes over time, encompassing both the value and trend of the change. The velocity correlation strength reflects the degree of correlation between the speeds of two drone nodes; a larger value indicates a closer speed relationship. Feature dispersion refers to the dispersion of deep correlation features in the three-dimensional coordinate system and measures the degree of feature dispersion. The feature screening parameter, determined by the maximum feature dispersion values ​​in the three coordinate systems, is used to screen key features. Key features are deep correlation features obtained after filtering using the feature screening parameters and are those with a significant impact on communication. The signal attenuation model parameters are pre-set parameters used to calculate compensation components, including the distance attenuation coefficient and the motion interference coefficient. The distance compensation component is the product of the distance change rate of the key feature and the distance attenuation coefficient, used to compensate for the effects of distance changes. The motion compensation component is the product of the velocity correlation strength of the key feature and the motion interference coefficient, used to compensate for interference caused by motion. The protocol layer compensation coefficient is a coefficient obtained by weighted superposition of the distance compensation component and the motion compensation component, and is used to correct communication protocol related features.

[0042] In the embodiment of the present application, step S1031 is first performed based on the spatial topology network of the communication environment model. A feature extraction algorithm based on association rules is then used to traverse all drone nodes and identify each pair of node combinations. The rate of change of the distance between any two nodes is then calculated by calculating the ratio of the distance difference between the two nodes at consecutive time points to the time interval. The velocity correlation strength is then obtained by analyzing the correlation between the velocity vectors of the two nodes. These two data sets are then used together as deep correlation features reflecting the dynamic relationship between the nodes. For example, in a network containing drones A, B, and C, it is calculated that the distance between A and B decreases by 0.3 meters per second, with a velocity correlation strength of 0.7. The distance between A and C increases by 0.5 meters per second, with a velocity correlation strength of 0.4.

[0043] Secondly, in step S1032, based on the deep correlation features, a standard deviation algorithm is used to calculate the distribution dispersion of these features in the three three-dimensional coordinate directions of x, y, and z, that is, the degree of feature dispersion. Each deep correlation feature is first decomposed into three coordinate directions, and then the standard deviation of all eigenvalues ​​in each direction is calculated separately. The larger the standard deviation, the higher the degree of dispersion. Then, the largest value of the feature dispersion in these three directions is selected as the feature screening parameter. For example, if the calculated standard deviation in the x direction is 2.2, the y direction is 2.8, and the z direction is 2.5, then 2.8 is determined as the feature screening parameter.

[0044] Next, in step S1033, the deep-level correlation features are screened using the feature screening parameters. The distance change rate and velocity correlation strength of each deep-level correlation feature are examined one by one. Only when a deep-level correlation feature's distance change rate is greater than the feature screening parameter and its velocity correlation strength is also greater than the feature screening parameter is it retained as a key feature. For example, if the feature screening parameter is 2.8, and a node pair has a distance change rate of 3.0 and a velocity correlation strength of 3.2, it meets the criteria and becomes a key feature. Features with a distance change rate of 2.7 or a velocity correlation strength of 2.6 are excluded.

[0045] Then, the filtered key features are processed through steps S1034 and S1035. First, the signal attenuation model preset in the system is called to obtain the distance attenuation coefficient. Then, the distance change rate of the key feature is multiplied by the distance attenuation coefficient. The result is the distance compensation component, where the distance attenuation coefficient is a fixed value pre-set according to the law that the signal attenuates with distance when propagating in space. For example, if the distance change rate of the key feature is 3.0 and the distance attenuation coefficient is 0.5, then the distance compensation component is 3.0×0.5=1.5; similarly, the motion interference coefficient is obtained from the preset signal attenuation model, and the speed correlation strength of the key feature is multiplied by the motion interference coefficient to obtain the motion compensation component. Wherein, the motion interference coefficient is a fixed value pre-set according to the degree of interference of the motion state on the signal. For example, if the speed correlation strength is 3.2 and the motion interference coefficient is 0.4, then the motion compensation component is 3.2×0.4=1.28.

[0046] Finally, in step S1036, the distance compensation component and the motion compensation component are combined. A weighted superposition calculation is performed on the two components based on the different weights of the impact of distance change and motion interference on the communication protocol. The weight values ​​are determined in advance based on a large amount of experimental data. Specifically, the distance compensation component is multiplied by its corresponding weight, and then the motion compensation component is multiplied by its corresponding weight. The result is the protocol layer compensation coefficient. For example, if the distance compensation component has a weight of 0.6 and the motion compensation component has a weight of 0.4, the protocol layer compensation coefficient is 1.5 × 0.6 + 1.28 × 0.4 = 0.9 + 0.512 = 1.412.

[0047] In practical applications, in a certain UAV joint monitoring scenario, the spatial topology network of the communication environment model includes UAVs P, Q, R, and S. First, deep correlation features are extracted from the network, such as the distance change rate between P and Q is 3.2 meters per second, the speed correlation strength is 3.0, the distance change rate between Q and R is 2.9 meters per second, the speed correlation strength is 3.5, the distance change rate between R and S is 3.6 meters per second, the speed correlation strength is 3.3, and the distance change rate between S and P is 2.7 meters per second, the speed correlation strength is 2.8; then the feature dispersion of these features in the x, y, and z directions is calculated, which are 3.0, 3.3, and 2.9 respectively, and 3.3 is selected as the feature screening parameter; then the features with distance change rate and speed correlation strength greater than 3.3 are screened out, namely R and S, The corresponding distance change rate is 3.6>3.3, and the speed correlation strength is 3.3 equal to 3.3, which is inconsistent. Assuming that the distance change rate of Q and S in the supplementary data is 3.4 and the speed correlation strength is 3.5, it meets the conditions and becomes a key feature; then multiply the distance change rate of the key feature 3.4 by the preset distance attenuation coefficient 0.6 to obtain the distance compensation component 3.4×0.6=2.04, and multiply the speed correlation strength 3.5 by the preset motion interference coefficient 0.5 to obtain the motion compensation component 3.5×0.5=1.75; finally, these two components are weightedly superimposed, the distance compensation component weight is 0.5, the motion compensation component weight is 0.5, and the protocol layer compensation coefficient is 2.04×0.5+1.75×0.5=1.02+0.875=1.895.

[0048] The overall solution in S103 extracts deep-level correlation features from the communication environment model, combines them with their dimensional distribution characteristics to identify key features, and then fuses these features with signal attenuation model parameters to generate protocol-level compensation coefficients. This process accurately captures features that significantly impact communication. By generating compensation coefficients, it effectively mitigates the effects of distance variations and motion interference on the communication protocol, improving the accuracy and reliability of the communication protocol features and making the communication system more adaptable to the dynamically changing drone operating environment.

[0049] S104. Extracting spatiotemporal variation characteristics of the communication network based on the positional relationship characteristics output by the closed-loop geometric modeling, predicting the communication interruption path based on the spatiotemporal variation characteristics, extracting communication protocol characteristics based on the protocol feature correction parameters, protocol layer compensation coefficients, and communication interruption path, and evaluating the communication interruption risk probability, so as to generate a dynamic selection strategy for backup communication nodes at the communication protocol layer.

[0050] Optionally, step S104 may specifically include the following steps: S1041, obtaining the coordinates of all boundary vertices in the final geometric model output by the closed-loop geometric modeling, and calculating the distances between adjacent vertices to form a positional relationship feature sequence; S1042: Analyze the position relationship feature sequence, extract the distance fluctuation range in each time window as an amplitude feature, and calculate the number of distance zero-crossing changes per unit time as a frequency feature, and combine the amplitude feature and the frequency feature to form a spatiotemporal variation feature; S1043. Detect boundary line segments in the spatiotemporal variation characteristics that satisfy a continuous increase in amplitude and a frequency exceeding a preset frequency threshold, determine the extension direction vectors of these line segments as candidate directions of the communication interruption path, perform cluster analysis on the candidate directions of the communication interruption path, and take the direction of the largest cluster center as the communication interruption path.

[0051] S1044. Combining the module length of the protocol feature correction parameter, the protocol layer compensation coefficient, and the azimuth and elevation angles of the communication interruption path into a communication protocol feature vector, inputting the communication protocol feature vector into a pre-trained risk probability calculation function, and outputting the communication interruption risk probability of each UAV node relative to the base station; S1045. Sort the communication interruption risk probabilities of all drone nodes in ascending order, and select the top N nodes with the lowest communication interruption risk probabilities to form a backup communication node set; S1046. Continuously monitor the quality parameter of the primary communication link. When the quality parameter is lower than a first quality threshold, switch the communication routing node to a node with the highest priority in the set of backup nodes. Detect the quality parameter of the new communication link within a fixed verification time window after the switch. If the quality parameter of the new communication link does not recover to a second quality threshold, remove the current node from the set of backup nodes and switch back to the node with the highest priority among the remaining nodes. S1047: When there is no available node in the backup node set, re-execute the interruption risk probability calculation and generate a new backup node set.

[0052] In the above scheme, the positional relationship feature sequence is a sequence of distances between adjacent boundary vertices in the final geometric model output by closed-loop geometric modeling, reflecting changes in the positional relationships between vertices. The spatiotemporal variation feature is a combination of amplitude and frequency features within each time window. The amplitude feature is the range of distance fluctuations, reflecting the magnitude of the variation, while the frequency feature is the number of times the distance crosses zero per unit time, reflecting the frequency of the variation. A communication interruption path is a directional path along which communication interruption may occur, determined by the extension direction of boundary segments that meet specific conditions. The communication protocol feature vector is a vector composed of the modulus of the protocol feature correction parameter, the protocol layer compensation coefficient, and the azimuth and elevation angles of the communication interruption path. It is used to assess communication interruption risk. The risk probability calculation function is a pre-trained function for calculating communication interruption risk probability. The communication interruption risk probability measures the probability of communication interruption for each drone node relative to the base station. The backup communication node set is a set of the top N nodes with the lowest communication interruption risk probability, serving as a replacement in the event of a primary link failure. The quality parameter of the primary communication link is an indicator that measures the communication quality of the primary communication link. The first and second quality thresholds are standard values ​​for determining whether the communication link quality meets the quality standards. The fixed verification time window is a fixed period of time for detecting the quality of the new link after switching nodes.

[0053] In an embodiment of the present application, first, in step S1041, the three-dimensional coordinates of all boundary vertices are extracted from the final geometric model output by the closed-loop geometric modeling, and the Euclidean distance calculation formula is used to calculate the straight-line distance between each pair of adjacent boundary vertices in turn. These distances are arranged in the order of vertex connection to form a positional relationship feature sequence. For example, the boundary vertex coordinates of the final geometric model are (1,1,1), (3,3,3), and (5,5,5), and the adjacent vertex distances are 3.46 and 3.46, respectively, forming the sequence [3.46,3.46]. Step S1042 uses a sliding window technique to divide the sequence into multiple time windows based on the sequence, analyzes the distance value in each window, and obtains the amplitude feature by subtracting the minimum value from the maximum value of the distance in the window. At the same time, the number of times the distance value changes from greater than zero to less than zero or from less than zero to greater than zero in a unit time is counted as the frequency feature, and the two are combined into a spatiotemporal change feature. For example, if the maximum value of the distance in a window is 4.0 and the minimum value is 2.0, the amplitude feature is 2.0, and there are 2 zero-crossing changes in a unit time, the frequency feature is 2, and the combination is a spatiotemporal change feature of (2.0, 2).

[0054] Next, in step S1043, based on the spatiotemporal variation characteristics, a trend analysis algorithm is used to screen out boundary line segments whose amplitude characteristics continuously increase and whose frequency characteristics exceed a preset frequency threshold. A directional vector calculation method is then used to determine the extended directional vectors of these line segments as candidate directions for the communication interruption path. The candidate directions are then clustered using the K-means clustering algorithm, and the cluster center direction containing the largest number of candidate directions is determined as the communication interruption path. For example, there are three candidate directional vectors: (0.6, 0.5, 0.3), (0.7, 0.4, 0.2), and (0.2, 0.3, 0.8). After clustering, the largest cluster center direction is (0.65, 0.45, 0.25), which is the communication interruption path.

[0055] Then, if Figure 2 As shown, in step S1044, the modulus of the protocol feature correction parameter vector is calculated, and the protocol layer compensation coefficient, as well as the azimuth and elevation angles of the communication interruption path, are obtained. These four data points are sequentially combined into a communication protocol feature vector. This is then input into a risk probability calculation function trained based on historical data to obtain the communication interruption risk probability of each drone node relative to the base station. For example, if the vector is (3.0, 2.2, 45°, 30°), the risk probability of a particular node after inputting the function is 0.15. Next, in step S1045, the risk probabilities of all nodes are sorted in ascending order from smallest to largest, and the top N nodes are selected to form the backup communication node set. For example, if the sorted probabilities are 0.1, 0.15, 0.2, and 0.3, the top two nodes are selected.

[0056] Finally, through steps S1046 and S1047, the quality parameter of the primary communication link, i.e., signal clarity, is continuously tracked. When the quality parameter falls below the first quality threshold, the communication route is switched to the node with the highest priority (i.e., the lowest probability) in the backup set. The quality of the new link is then tested within a set fixed verification time window. If the second quality threshold is not reached, the node is removed from the backup set and switched to the node with the highest priority among the remaining nodes. If there are no available nodes in the backup set, the risk probability calculation step of step S1044 is re-executed to generate a new set of backup communication nodes. For example, if the quality parameter of the primary communication link is 55, which is lower than the first quality threshold of 60, the node is switched to a node with a probability of 0.1. If the new link quality of 65 does not reach the second quality threshold of 70 within 8 seconds, the node is removed and the node is switched to a node with a probability of 0.15. If this node also fails, the new set is recalculated and generated. In practical applications, in a certain UAV collaborative reconnaissance scenario, the final geometric model boundary vertex coordinates output by closed-loop geometric modeling are (2,2,2), (4,4,4), (6,6,6), (2,2,2), and the calculated distances between adjacent vertices are 3.46, 3.46, and 5.66, respectively, forming a position relationship feature sequence [3.46,3.46,5.66]. Then, a sliding window is used to divide the time window. The maximum distance in a certain window is 6.0, the minimum distance is 3.0, the amplitude feature is 3.0, the unit time crosses zero 3 times, and the frequency feature is 3, which is combined into a spatiotemporal variation feature of (3.0,3). According to the spatiotemporal variation feature, the boundary line segments with continuously increasing amplitude and frequency exceeding the threshold 2 are screened out, and their extension direction vectors are (0.5,0.5,0.5) and (0.6,0.6,0.6). After K -means clustering obtains the direction of the maximum cluster center (0.55, 0.55, 0.55) as the communication interruption path; the protocol feature correction parameter modulus is calculated as 4.0, the protocol layer compensation coefficient is 2.5, the interruption path azimuth angle is 60°, and the pitch angle is 30°, which are combined into a feature vector (4.0, 2.5, 60°, 30°). The pre-training function is input to obtain the interruption risk probabilities of drones H, I, J, and K as 0.12, 0.18, 0.25, and 0.3, respectively. After sorting, S1045 selects the first two to form the backup set {H, I}; finally, the main link quality parameter 58 is monitored to be lower than the first threshold 60, and the link is switched to H. Within 8 seconds, the new link quality 68 does not reach the second threshold 70. After removing H, the link is switched to I, and the new link quality 72 meets the standard. If I also fails, S1047 recalculates and generates a new backup set.

[0057] The S104 solution, based on closed-loop geometric modeling, extracts spatial and temporal variation characteristics based on positional relationships, accurately predicts communication interruption paths, and combines relevant parameters to assess risk probability and generate a dynamic backup node selection strategy. By monitoring the quality of the primary link in real time and dynamically switching to backup nodes, it can promptly address communication interruption risks, ensuring communication continuity and stability, and effectively improving the anti-interference capability and reliability of the UAV communication network in complex and dynamic scenarios.

[0058] The following is a complete example for steps 101 to 104. Figure 3As shown in the figure, in a drone collaborative operation scenario in a certain area, there is a base station P and a group of drones, namely drones A, B, C, and D. First, the spatial position set of base station P and drones A, B, C, and D is obtained, and their three-dimensional coordinates are obtained. The real-time speed of each location is also recorded, such as drone A's speed of 4 m / s southeast and drone B's speed of 3 m / s northeast. Acceleration information is also recorded, such as drone C's acceleration of 1.5 m / s² southward. Based on this spatial position set, a spatial topology network is established, including base station P and each drone, and the real-time speeds are annotated. Dynamic connection relationships are then constructed based on the distance changes and speed differences between adjacent nodes. Finally, acceleration information is integrated to form a communication environment model that can be updated in real time. Next, the position coordinates of drones A, B, C, and D are input into the spatial mapping rule. The coordinates of the outermost layers, A, C, and D, are connected to form a triangular virtual spatial shape. The arithmetic mean of the coordinates of all vertices of this shape is calculated to obtain the geometric center coordinates (4, 5, 6). The coordinates of the center of base station P's communication range are (5, 5, 6), and the spatial offset between the two is -1 in the x-direction. Then, at a preset interval of 2 seconds, the instantaneous velocity change and displacement compensation components of each drone are calculated. After superposition, the overall displacement compensation vector is obtained as (0.9, 0, 0). Vector subtraction is performed between the spatial offset and the overall displacement compensation vector to obtain the protocol feature correction parameter (-1.9, 0, 0). The vertex positions of the virtual spatial shape are adjusted based on these parameters. After generating a new virtual spatial shape, the closed-loop verification offset between its geometric center and the center of the communication range is calculated to be (-0.8, 0, 0). This value is less than the preset threshold of 1, so the current virtual spatial shape is output as the final geometric model. Next, deep correlation features were extracted from the spatial topology network of the communication environment model. For example, the distance change rate between A and B was 3.8 meters per second, with a velocity correlation strength of 3.6; the distance change rate between B and C was 4.2 meters per second, with a velocity correlation strength of 3.9. The feature dispersion of these features in the x, y, and z directions was calculated to be 3.1, 3.7, and 3.3, respectively. A feature screening parameter of 3.7 was selected, and features with both a distance change rate and a velocity correlation strength greater than 3.7 were selected, namely, the correlation features between B and C, as key features. The distance change rate of this key feature, 4.2, was multiplied by a preset distance attenuation coefficient of 0.6 to obtain a distance compensation component of 2.52. The velocity correlation strength of 3.9 was multiplied by a preset motion interference coefficient of 0.5 to obtain a motion compensation component of 1.95. These two components were then weighted and superimposed, with weights of 0.5 and 0.5, respectively, to generate a protocol-layer compensation coefficient of 2.235. Finally, the coordinates of the boundary vertices in the final geometric model output by the closed-loop geometric modeling are obtained. The distances between adjacent vertices are calculated to form a positional relationship feature sequence. This sequence is analyzed to obtain spatiotemporal variation features. From these features, eligible boundary segments are screened, candidate directions of communication interruption paths are determined and clustered, resulting in the communication interruption paths. The module length of the protocol feature correction parameters, the protocol layer compensation coefficient of 2.235, the azimuth angle of 35°, and the pitch angle of 40° of the communication interruption path are combined into a communication protocol feature vector. This vector is then fed into a pre-trained risk probability calculation function, yielding communication interruption risk probabilities of 0.12, 0.18, 0.25, and 0.21 for drones A, B, C, and D, respectively. The first two are selected to form the backup communication node set {A, B}. The quality parameters of the primary communication link are continuously monitored. When the quality falls below the first quality threshold of 55, the link switches to A. The new link quality is then tested within a fixed verification window of 10 seconds. If the quality of the new link does not recover to the second quality threshold of 65, A is removed and the link switches to B. If B also fails, a new backup node set is recalculated.

[0059] Figure 4 This is a structural diagram of a specific implementation of an artificial intelligence-based drone communication protocol feature extraction system provided in an embodiment of the present application, with reference to Figure 4 , the system may include: An acquisition module 41 is configured to acquire a spatial position set of the base station and the drone swarm, synchronously collect real-time velocity and acceleration information of each position point in the spatial position set, and construct a communication environment model based on the spatial position set, real-time velocity and acceleration information; a compensation module 42 for mapping the set of spatial positions to generate a virtual spatial shape, calculating a spatial offset between a geometric center point of the virtual spatial shape and a center point of a communication range, performing motion compensation on the spatial offset in combination with the acceleration information to generate protocol feature correction parameters, and adjusting shape parameters of the virtual spatial shape using feedback from the protocol feature correction parameters to form a closed-loop geometric modeling; A screening module 43 is configured to extract deep correlation features of the communication environment model, generate feature screening parameters based on the dimensional distribution characteristics of the communication environment model, screen key features of the deep correlation features based on the feature screening parameters, and fuse the screened key features with preset signal attenuation model parameters to generate protocol layer compensation coefficients; A generation module 44 is used to extract the spatiotemporal variation characteristics of the communication network based on the positional relationship characteristics output by the closed-loop geometric modeling, predict the communication interruption path based on the spatiotemporal variation characteristics, extract the communication protocol characteristics based on the protocol characteristic correction parameters, protocol layer compensation coefficients and communication interruption paths, and evaluate the communication interruption risk probability, so as to generate a dynamic selection strategy for backup communication nodes in the communication protocol layer.

[0060] The artificial intelligence-based drone communication protocol feature extraction system of the embodiment of the present application is used to implement the aforementioned artificial intelligence-based drone communication protocol feature extraction method. Therefore, the specific implementation method of the artificial intelligence-based drone communication protocol feature extraction system can be seen in the embodiment part of the artificial intelligence-based drone communication protocol feature extraction method in the previous text. Its specific implementation method can refer to the description of the corresponding embodiments of each part, which will not be repeated here.

[0061] The present application also provides an electronic device, comprising: a memory for storing a computer program; and a processor for implementing the steps of any of the above-mentioned artificial intelligence-based drone communication protocol feature extraction methods when executing the computer program.

[0062] The present application also provides a computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, the steps of any of the above-mentioned artificial intelligence-based drone communication protocol feature extraction methods are implemented.

[0063] In an exemplary embodiment, the computer-readable storage medium may include, but is not limited to, various media capable of storing computer programs, such as a USB flash drive, a read-only memory, a random access memory, a mobile hard disk, a magnetic disk, or an optical disk.

[0064] An embodiment of the present invention also provides a computer program product, which includes a computer program. When the computer program is executed by a processor, it implements the steps of any of the above-mentioned artificial intelligence-based drone communication protocol feature extraction method embodiments.

[0065] Professionals may further appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the above description has generally described the components and steps of each example according to their functions. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professionals and technicians may use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of the present invention.

[0066] The above is a detailed introduction to the artificial intelligence-based drone communication protocol feature extraction method, system, electronic device, and storage medium provided by this application. This article uses specific examples to illustrate the principles and implementation methods of this application. The description of the above embodiments is only used to help understand the method and core ideas of this application. It should be pointed out that for ordinary technicians in this technical field, without departing from the principles of this application, several improvements and modifications can be made to this application, and these improvements and modifications also fall within the scope of protection of this application.

Claims

1. A method for extracting features of UAV communication protocols based on artificial intelligence, characterized in that: include: Acquire a spatial position set of the base station and the drone swarm, synchronously collect real-time velocity and acceleration information of each position point in the spatial position set, and construct a communication environment model based on the spatial position set, real-time velocity and acceleration information; Mapping the set of spatial positions to generate a virtual spatial shape, calculating a spatial offset between a geometric center point of the virtual spatial shape and a center point of a communication range, performing motion compensation on the spatial offset in combination with the acceleration information to generate a protocol feature correction parameter, and adjusting shape parameters of the virtual spatial shape through feedback of the protocol feature correction parameter to form a closed-loop geometric modeling; Extracting deep correlation features of the communication environment model, generating feature screening parameters based on the dimensional distribution characteristics of the communication environment model, screening key features of the deep correlation features based on the feature screening parameters, and fusing the screened key features with preset signal attenuation model parameters to generate protocol layer compensation coefficients; Based on the positional relationship features output by the closed-loop geometric modeling, the spatiotemporal variation features of the communication network are extracted, and the communication interruption path is predicted based on the spatiotemporal variation features. Based on the protocol feature correction parameters, the protocol layer compensation coefficient and the communication interruption path, the communication protocol features are extracted and the communication interruption risk probability is evaluated to generate a dynamic selection strategy for backup communication nodes at the communication protocol layer.

2. The method according to claim 1, characterized in that The extracting of communication protocol features and evaluating the communication interruption risk probability based on the protocol feature correction parameter, the protocol layer compensation coefficient, and the communication interruption path to generate a dynamic selection strategy for a backup communication node at the communication protocol layer includes: The module length of the protocol feature correction parameter, the protocol layer compensation coefficient, the azimuth angle and the pitch angle of the communication interruption path are combined into a communication protocol feature vector, and the communication protocol feature vector is input into the pre-trained risk probability calculation function to output the communication interruption risk probability of each drone node relative to the base station; Sort the communication interruption risk probabilities of all drone nodes in ascending order, and select the top N nodes with the lowest communication interruption risk probability to form the backup communication node set; Continuously monitor the quality parameters of the primary communication link. When the quality parameters are lower than a first quality threshold, switch the communication routing node to the node with the highest priority in the backup node set. Detect the quality parameters of the new communication link within a fixed verification time window after the switch. If the quality parameters of the new communication link do not recover to a second quality threshold, remove the current node from the backup node set and switch back to the node with the highest priority among the remaining nodes. When there is no available node in the standby node set, the interruption risk probability calculation is re-executed to generate a new standby node set.

3. The method according to claim 1, characterized in that The adjusting of the shape parameters of the virtual space shape by feedback of the protocol feature correction parameters to form a closed-loop geometric modeling includes: Determining a vertex movement direction and a movement distance of a virtual space shape based on the protocol feature correction parameter; performing a synchronous translation operation on all vertex coordinates of the virtual space shape according to the vertex movement direction and movement distance, and reconnecting the translated vertex coordinates to generate a new virtual space shape; Calculating the coordinates of the geometric center point of the new virtual space shape, and measuring the difference between the coordinates of the geometric center point and the coordinates of the center point of the base station communication range as a closed-loop verification offset; When the closed-loop verification offset is less than a preset threshold, the current virtual space shape is output as the final geometric model. When the closed-loop verification offset is greater than or equal to the preset threshold, the closed-loop verification offset is input as a new space offset into the motion compensation process to regenerate the protocol feature correction parameters.

4. The method according to claim 1, wherein The extracting of spatiotemporal variation characteristics of the communication network based on the positional relationship characteristics output by the closed-loop geometric modeling, and predicting the communication interruption path according to the spatiotemporal variation characteristics, includes: Obtain the coordinates of all boundary vertices in the final geometric model output by closed-loop geometric modeling, and calculate the distances between adjacent vertices to form a positional relationship feature sequence; Analyze the position relationship feature sequence, extract the distance fluctuation range in each time window as the amplitude feature, and calculate the number of distance zero-crossing changes per unit time as the frequency feature, and combine the amplitude feature and the frequency feature to form the spatiotemporal change feature; In the spatiotemporal variation characteristics, boundary line segments that satisfy the requirement of continuously increasing amplitude and a frequency exceeding a preset frequency threshold are detected, and the extension direction vectors of these line segments are determined as candidate directions of the communication interruption path. Cluster analysis is performed on the candidate directions of the communication interruption path, and the direction of the maximum cluster center is taken as the communication interruption path.

5. The method according to claim 1, wherein Mapping the spatial position set to generate a virtual spatial shape, calculating a spatial offset between a geometric center point of the virtual spatial shape and a center point of a communication range, and performing motion compensation on the spatial offset in combination with the acceleration information to generate a protocol feature correction parameter, including: Inputting the position coordinates of the drones in the spatial position set into the spatial mapping rule, and forming a virtual spatial shape by connecting the position coordinates of the outermost drones; Calculating the arithmetic mean of all vertex coordinates of the virtual space shape as the coordinates of the geometric center point, and simultaneously measuring the spatial offset between the coordinates of the geometric center point and the coordinates of the center point of the base station communication range; Multiplying the acceleration information of each drone node by a preset time interval to obtain an instantaneous velocity change, multiplying the instantaneous velocity change by the same time interval to obtain a displacement compensation component, and performing vector superposition on the displacement compensation components of all drone nodes to generate an overall displacement compensation vector; A vector subtraction operation is performed on the spatial offset and the overall displacement compensation vector to obtain a protocol feature correction parameter.

6. The method according to claim 1, characterized in that The extracting of deep correlation features of the communication environment model, generating feature screening parameters according to the dimensional distribution characteristics of the communication environment model, screening key features of the deep correlation features based on the feature screening parameters, and fusing the screened key features with preset signal attenuation model parameters to generate protocol layer compensation coefficients includes: The distance change rate and speed correlation strength between all drone node pairs are extracted from the spatial topology network of the communication environment model as deep correlation features; Calculate the feature dispersion of the deep correlation feature in the three-dimensional coordinate direction, and take the maximum dispersion value in the three directions as the feature screening parameter; Screening out deep correlation features whose distance change rate is greater than the feature screening parameter and whose velocity correlation strength is greater than the feature screening parameter as key features; Multiplying the distance change rate of the key feature by a preset distance attenuation coefficient to obtain a distance compensation component; Multiplying the velocity correlation strength of the key feature by a preset motion interference coefficient to obtain a motion compensation component; The distance compensation component and the motion compensation component are weightedly superimposed to generate a protocol layer compensation coefficient.

7. The method according to claim 1, characterized in that The acquiring of a spatial position set of the base station and the drone swarm, synchronously collecting real-time velocity and acceleration information of each position point in the spatial position set, and constructing a communication environment model based on the spatial position set, the real-time velocity and acceleration information, includes: Obtain the spatial position set of the base station and all individual drones in the drone group, and synchronously record the real-time speed and acceleration information corresponding to each position point in the spatial position set; Establishing a spatial topology network including base station nodes and drone nodes based on the spatial position set, and marking the real-time speeds of the base station nodes and drone nodes at corresponding positions in the spatial topology network; Building a dynamic connection relationship in the spatial topology network according to the spatial position distance change and real-time speed difference between adjacent nodes; The acceleration information is integrated into the dynamic connection relationship as a motion state variable to form a communication environment model that can be updated in real time.

8. An artificial intelligence-based UAV communication protocol feature extraction system, characterized in that: include: An acquisition module is configured to acquire a spatial position set of the base station and the drone swarm, synchronously collect real-time velocity and acceleration information of each position point in the spatial position set, and construct a communication environment model based on the spatial position set, real-time velocity and acceleration information; a compensation module, configured to map the set of spatial positions to generate a virtual spatial shape, calculate a spatial offset between a geometric center point of the virtual spatial shape and a center point of a communication range, perform motion compensation on the spatial offset in combination with the acceleration information to generate protocol feature correction parameters, and adjust shape parameters of the virtual spatial shape using feedback from the protocol feature correction parameters to form a closed-loop geometric modeling; a screening module, configured to extract deep correlation features of the communication environment model, generate feature screening parameters based on the dimensional distribution characteristics of the communication environment model, screen key features of the deep correlation features based on the feature screening parameters, and fuse the screened key features with preset signal attenuation model parameters to generate protocol layer compensation coefficients; A generation module is used to extract the spatiotemporal variation characteristics of the communication network based on the positional relationship characteristics output by the closed-loop geometric modeling, predict the communication interruption path according to the spatiotemporal variation characteristics, extract the communication protocol characteristics based on the protocol characteristic correction parameters, protocol layer compensation coefficients and communication interruption path, and evaluate the communication interruption risk probability, so as to generate a dynamic selection strategy for backup communication nodes in the communication protocol layer.

9. An electronic device, characterized in that: include: memory for storing computer programs; A processor, configured to implement the steps of the method for extracting features of a drone communication protocol based on artificial intelligence as described in any one of claims 1 to 7 when executing the computer program.

10. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a computer program, which, when executed by a processor, can implement the artificial intelligence-based drone communication protocol feature extraction method according to any one of claims 1 to 7.

Images