Unmanned aerial vehicle cluster interaction method and system based on ad hoc network

Abstract

The invention discloses an unmanned aerial vehicle cluster interaction method and system based on an ad hoc network, and belongs to the technical field of unmanned aerial vehicle communication and cluster control. The method comprises the following steps: acquiring positioning information, energy parameters and task types of an unmanned aerial vehicle group to generate a clustering node set; predicting a cluster moving direction based on the motion acceleration data of the clustering node set, and generating a dynamic path planning instruction; allocating a main communication channel and a standby channel according to the dynamic path planning instruction, and generating a channel allocation result; monitoring a signal quality parameter of the main communication channel according to a channel allocation result, and triggering a link switching instruction when interference is detected; and constructing an interference thermodynamic diagram based on the topological relation of the clustering node set, and generating a power adjustment instruction. According to the invention, through dynamic clustering, mobile prediction, adaptive channel management, interference sensing switching and topology power control, the communication stability, reliability and resource efficiency of the unmanned aerial vehicle cluster in a complex environment are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of UAV communication, and particularly to a method and system for UAV cluster interaction based on an ad hoc network. Background Art

[0002] A UAV cluster system, which consists of multiple UAVs with autonomous capabilities, is being increasingly applied to scenarios that require wide-area coverage, distributed cooperation, or operation in high-risk environments. In these applications, each UAV unit within the cluster needs to cooperate closely, share perception information, and coordinate flight trajectories and mission actions. Since the mission environment often does not have the support of a fixed ground communication network, or relying on the ground network will introduce latency and single-point failure risks, the related art generally adopts the method of directly establishing wireless communication links between UAVs to form a dynamic and self-organizing aerial network to undertake the necessary information transmission function within the cluster.

[0003] The smoothness and final effect of the cluster executing tasks as a whole are sometimes restricted. It is manifested that in some operation stages or specific environments, the transmission process of the information used for coordination between nodes does not always achieve the stability and timeliness required by the task. This situation may be more obvious when the cluster is performing rapid maneuvers, formation changes, or operating in a complex physical space, which may in turn affect the accuracy of the cluster's collaborative actions and the quality of task completion. Summary of the Invention

[0004] To solve the above problems, the present invention provides a method and system for UAV cluster interaction based on an ad hoc network. By adopting a dynamic clustering management network structure, combining path planning and channel pre-allocation based on motion prediction, implementing an interference-aware link monitoring and switching mechanism, and using an interference heat map based on topological relationships for power control and topological weight adjustment, the adaptive ability, anti-interference performance, and overall network efficiency of UAV cluster communication are significantly improved.

[0005] The above objectives can be achieved through the following solutions: A method for UAV cluster interaction based on an ad hoc network includes obtaining the positioning information, remaining energy parameters, and real-time task types of the UAV group to generate a set of clustering nodes; extracting the motion acceleration data of the UAV cluster based on the set of clustering nodes, predicting the moving direction of the cluster, and generating a dynamic path planning instruction; allocating a primary communication channel according to the dynamic path planning instruction to generate a channel allocation result; monitoring the signal quality parameters of the primary communication channel according to the channel allocation result, and triggering a link switching instruction when interference is detected; constructing an interference heat map according to the topological relationship of the set of clustering nodes to generate a power adjustment instruction.

[0006] Optionally, the generating of the cluster node set includes: extracting the correlation degree parameter between the positioning information and the real-time task type, and screening out a preset candidate node set that meets the cluster head conditions; determining the main cluster head node from the candidate node set based on the remaining energy parameter and the preset movement stability coefficient; dividing the member range of the cluster node set according to the communication radius of the main cluster head node and the preset signal strength threshold, and generating the cluster node set.

[0007] Optionally, the generating of the dynamic path planning instruction includes: collecting the historical movement data of the UAV cluster, and constructing a movement trend prediction model; collecting the acceleration vector sequence of the cluster node set within a preset time window, and inputting it into the movement trend prediction model to generate a predicted movement path; collecting the communication load data of the cluster node set, and generating channel resource pre-allocation parameters in combination with the predicted movement path; generating a dynamic path planning instruction according to the channel resource pre-allocation parameters and the predicted movement path, and sending it to the main cluster head node.

[0008] Optionally, the triggering of the link switching instruction includes: continuously monitoring the bit error rate and signal attenuation rate of the main communication channel, and generating an interference level evaluation parameter; matching the preset frequency hopping sequence library according to the interference level evaluation parameter, selecting a target frequency hopping mode, and generating a link switching instruction and triggering the power compensation instruction of the cluster node set.

[0009] Optionally, the triggering of the power compensation instruction includes: determining whether the communication frequency band corresponding to the target frequency hopping mode overlaps with an external interference source, and generating a frequency band conflict identifier according to the determination result; adjusting the power difference of the co-frequency nodes in the cluster node set and reducing the non-critical task bandwidth occupancy ratio according to the frequency band conflict identifier.

[0010] Optionally, the constructing of the interference heat map includes: obtaining the real-time three-dimensional coordinates of the cluster node set, and calculating the signal coverage overlapping area between adjacent nodes; generating an interference hot spot distribution map according to the signal coverage overlapping area and the external spectrum scanning data; dynamically adjusting the topological connection weight of the cluster node set based on the interference hot spot distribution map.

[0011] Optionally, the dynamically adjusting the topological connection weight of the cluster node set based on the interference hot spot distribution map includes: when it is detected that the signal coverage overlapping area of the cross-cluster node exceeds a preset threshold, generating an arbitration request instruction and sending it to the ground control station; receiving the global topological optimization instruction fed back by the ground control station, and updating the communication priority configuration of the cluster node set.

[0012] Optionally, the channel resource pre-allocation parameter includes: calculating the remaining available bandwidth resource according to the channel resource pre-allocation parameter; when the remaining available bandwidth resource is lower than a preset threshold, preferentially allocating additional channel time slots to the node corresponding to the power compensation instruction.

[0013] Optionally, adjusting the power difference of the co-frequency nodes in the cluster node set further includes: obtaining a historical frequency band conflict identifier, extracting the occurrence frequency and time period distribution data of the frequency band conflict event to obtain a conflict data set; updating the priority sorting rule of the frequency hopping sequence library according to the conflict data set.

[0014] Based on the same inventive concept, the present invention also provides an unmanned aerial vehicle cluster interaction system based on an ad-hoc network. The system includes: an information acquisition and clustering module, configured to acquire the positioning information, remaining energy parameter, and real-time task type of the unmanned aerial vehicle group, and generate a cluster node set; a movement prediction and planning module, configured to predict the cluster movement direction based on the movement acceleration data of the cluster node set, and generate a dynamic path planning instruction; a channel allocation module, configured to allocate a primary communication channel and activate a standby channel according to the dynamic path planning instruction, and generate a channel allocation result; a link monitoring and switching module, configured to monitor the signal quality parameter of the primary communication channel in the channel allocation result, and trigger a link switching instruction when interference is detected; a topology and power control module, configured to construct an interference heat map according to the topological relationship of the cluster node set, and generate a power adjustment instruction.

[0015] Compared with the prior art, the present invention has the following advantages: 1. Enhanced the adaptive ability of the network in a high-dynamic environment: By dynamically generating a cluster node set based on real-time information and combining the prediction of the cluster movement direction to generate a dynamic path planning instruction, the network topology management and resource planning can actively adapt to the high-speed movement and formation change of the unmanned aerial vehicle nodes, maintaining the stability and continuity of the communication link.

[0016] 2. Improved the anti-interference performance of cluster communication: Combining continuous monitoring of the signal quality of the primary communication channel, frequency hopping mode selection based on interference level evaluation, frequency band conflict detection and processing, and power adjustment based on the interference heat map can effectively avoid or suppress internal and external interference, ensuring communication reliability in a complex electromagnetic environment.

[0017] 3. Optimized the utilization efficiency of spectrum and energy resources: By embedding channel resource pre-allocation parameters in the dynamic path planning, optimizing the activation logic of the standby channel based on the predicted remaining available bandwidth, and performing refined power adjustment according to the interference heat map, unnecessary channel occupancy, conflicts, and energy consumption are reduced, improving the spectrum utilization rate and the effective operation time of the unmanned aerial vehicle.

[0018] 4. Enhanced the support ability for cluster collaborative tasks: A stable, reliable, and efficient internal interaction network provides a basic guarantee for executing complex cluster collaborative tasks. By dynamically adjusting the topological connection weights based on the interference heat map and configuring the communication priorities under global coordination, the real-time performance and reliability of critical task data transmission are further ensured.

[0019] Other features and advantages of the present invention will be described in the following specification, and, in part, will become apparent from the specification or will be understood by implementing the present invention. The objectives and other advantages of the present invention can be achieved and obtained through the structures pointed out in the specification, claims, and drawings. Brief Description of the Drawings

[0020] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0021] Figure 1 It is a schematic flowchart of a method for UAV cluster interaction based on an ad-hoc network according to an embodiment of the present invention.

[0022] Figure 2 It is a schematic diagram of the clustering result according to an embodiment of the present invention.

[0023] Figure 3 It is a link quality monitoring and handover threshold graph according to an embodiment of the present invention.

[0024] Figure 4 It is an interference heat map according to an embodiment of the present invention.

[0025] Figure 5 It is a schematic structural diagram of a UAV cluster interaction system based on an ad-hoc network according to an embodiment of the present invention. Detailed Embodiments

[0026] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the drawings in the embodiments of the present invention. Obviously, the described embodiments are some, but not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present invention.

[0027] Refer to Figure 1, an embodiment of the present invention provides a method for UAV cluster interaction based on an ad-hoc network. This method aims to solve the problems of unstable communication and low network efficiency in UAV clusters during mission execution due to factors such as high-speed node movement, external interference, and energy limitations. Through a series of dynamic adjustment and optimization steps, reliable and efficient internal cluster interaction is achieved.

[0028] The method of this embodiment specifically includes the following steps: Obtain the positioning information, remaining energy parameters, and real-time task types of the UAV swarm, and generate a set of clustering nodes; Specifically, obtain the positioning information of each UAV, which is the node position and motion state data, usually from the combination of the Global Navigation Satellite System (GNSS) and the Inertial Navigation System (INS). At the same time, collect the remaining energy parameters, which is an indicator reflecting the available energy of the node, such as the battery percentage. Receive the real-time task type, which is an identifier indicating the current role of the node. Based on this information, run a clustering algorithm, such as according to proximity or task requirements, divide the cluster into several clusters, and form a set of clustering nodes, which is a logical subset set containing cluster heads and member nodes, used to simplify network management.

[0029] Extract the motion acceleration data of the UAV cluster based on the set of clustering nodes, predict the moving direction of the cluster, and generate a dynamic path planning instruction; Specifically, collect the recent motion acceleration data of the nodes, which is a vector reflecting the speed change. Apply a moving trend prediction model, which is a model using historical motion data to predict future trajectories, such as the Kalman filter or the Recurrent Neural Network (RNN), to predict the moving direction of the cluster, which is the expected motion trend of the cluster or sub-cluster. Combine this prediction with task, obstacle avoidance, and other requirements to generate a dynamic path planning instruction, which is an adjustable flight and behavior control command, providing a pre-judgment for communication management.

[0030] Allocate the main communication channel according to the dynamic path planning instruction, and generate a channel allocation result; Specifically, analyze the implied future network changes and load requirements in the dynamic path planning instruction. Accordingly, allocate a suitable main communication channel for the communication links in the cluster, which is the current main transmission resource, such as a specific frequency point or time slot. According to the link importance or risk assessment, decide whether to activate the backup channel, which is an alternative transmission resource. Finally, form a channel allocation result, which is information specifying the current channel configuration of each node and issue it for execution.

[0031] According to the channel allocation result, monitor the signal quality parameters of the primary communication channel, and trigger a link switching instruction when interference is detected; Specifically, the node continuously monitors the signal quality parameters of the primary channel in use, which are indicators to measure the channel performance, such as Received Signal Strength Indicator (RSSI) or Bit Error Rate (BER). When these parameters deteriorate to a preset threshold due to interference, that is, any factor causing signal deterioration, it is determined that the primary channel is disturbed. At this time, a link switching instruction is triggered, which is a command to instruct the node to switch to the backup channel to maintain the communication quality.

[0032] Construct an interference heat map based on the topological relationship of the cluster node set, and generate a power adjustment instruction.

[0033] Specifically, utilize the topological relationship after clustering, which is the connection structure information between nodes, combine the real-time positions and transmission characteristics of the nodes, and calculate the signal coverage and potential overlap. Integrate internal and external interference information to construct a dynamic interference heat map, which is a spatial visualization map of interference intensity or risk. Based on the interference distribution and coverage analyzed from the heat map, generate a power adjustment instruction, which is a command to adjust the transmission power of the node, aiming to optimize the overall network coverage, that is, the effective communication range of the signal, and improve the communication quality and efficiency.

[0034] By integrating real-time information for dynamic clustering, using mobile prediction to guide path planning and channel allocation, implementing interference-aware link switching, and performing power control based on the interference heat map, this method can effectively improve the communication stability and performance of the UAV swarm ad-hoc network in complex environments.

[0035] Optionally, the generation of the cluster node set includes: Extract the correlation parameter between the positioning information and the real-time task type, and screen out a preset candidate node set that meets the cluster head conditions; Specifically, this step aims to preliminarily screen potential cluster managers. Analyze the positioning information and real-time task types of each UAV (Unmanned Aerial Vehicle) node obtained, and calculate the correlation parameter between the nodes , which is an index to quantify the correlation between node and node in terms of geographical proximity, task similarity, or cooperation requirements, etc. An exemplary calculation method is as follows: , where, represents the correlation parameter value between node and node ; is the Euclidean distance between nodes and nodes , and is a distance-related function, such as a function indicating that the closer the distance, the higher the correlation degree respectively represent the real-time task types of nodes and and nodes and respectively. is a task similarity function that evaluates the degree of association between two tasks and and are the weight coefficients corresponding to the distance factor and the task factor respectively. At the same time, according to the preset cluster head conditions, this is a set of criteria for screening nodes with management capabilities, such as requiring the remaining energy, computing power, or network connectivity of the nodes to meet specific requirements. Compare the calculated correlation degree parameter with the threshold, and make a judgment in combination with whether the node itself meets the cluster head conditions to screen out the candidate node set, which is a subset of UAVs that are initially qualified to be cluster heads

[0036] Determine the main cluster head node from the candidate node set based on the remaining energy parameter and the preset mobile stability coefficient Specifically, this step is to select the most suitable cluster head from the candidates. Further evaluate the remaining energy parameter and the mobile stability coefficient of each node in the candidate node set. The mobile stability coefficient is an index to measure the smoothness of the recent movement state of the UAV node, which can be obtained by analyzing the node movement data. The smaller the value, the more stable it is. When selecting the main cluster head, nodes with sufficient remaining energy and high mobile stability are usually given priority. A comprehensive evaluation value can be calculated for selection, for example: , where represents the comprehensive evaluation value of the candidate node, is the normalized remaining energy index, is the normalized mobile stability index, and the larger the value, the more stable it is, and are the weight coefficients corresponding to energy and stability respectively. Select the candidate node with the highest comprehensive evaluation value as the main cluster head node, which is the UAV node finally selected to be responsible for managing its cluster in a specific area or time period

[0037] Divide the member range of the sub-cluster node set according to the communication radius of the main cluster head node and the preset signal strength threshold to generate a sub-cluster node set

[0038] Specifically, once the main cluster head node is determined, it broadcasts its own identity information to the surrounding area. Other non-cluster head nodes or UAVs whose ownership has not been determined receive signals from one or more potential main cluster head nodes. Each ordinary UAV node needs to determine whether to join the management of a certain main cluster head . There are mainly two judgment bases. First, the distance between the node and the main cluster head must be within its effective communication radius . The communication radius is the maximum theoretical or actual distance that the signal of the main cluster head node can reliably cover. That is, the condition needs to be met: , , where the signal strength received by the node from the main cluster head must be higher than the preset signal strength threshold . The signal strength threshold is the minimum received signal power level required to ensure reliable communication. The received signal strength can be measured actually or estimated based on the wireless channel propagation model. Therefore, the second condition is: The ordinary UAV node usually selects the main cluster head node that simultaneously meets the above two conditions and has the strongest received signal strength to join its cluster. All UAV nodes that choose to join the same main cluster head together form the member scope of the cluster. The member scope is all member nodes that form a specific cluster and their spatial distribution areas. So far, the clustering process is completed, forming a structured set of clustered nodes, as shown. Figure 2 shown.

[0039] Exemplarily, consider a UAV cluster. U1, U2, and U3 are candidate cluster heads. By calculating the comprehensive evaluation value , U1 has the highest score and is selected as the main cluster head. Suppose the communication radius of U1 is 500 meters, and the signal strength threshold is -85 dBm. The ordinary UAV U4 is 400 meters away from U1 (less than ), and the signal strength received from U1 is measured or estimated to be -68 dBm (greater than ). Therefore, U4 meets the conditions for joining the U1 cluster. If U4 does not receive signals from other main cluster heads that meet the conditions and have stronger signals, then U4 joins the U1 cluster. All such UAVs that choose to join U1 together form the member scope of the U1 cluster.

[0040] Optionally, the generating of the dynamic path planning instruction includes: Collecting historical motion data of the UAV cluster to construct a moving trend prediction model; Specifically, this step aims to establish a basic model for prediction. First, collect and store the historical motion data accumulated by the UAV cluster during past operations. This data may include time series of the position p, velocity v, and acceleration a of each UAV. Based on this historical motion data, use a specific algorithm or technology to train offline or construct online a moving trend prediction model. For example, a Kalman Filter (KF) model based on state space can be used. In this model, the motion state of the UAV can be represented by a state vector For example: , wherein it contains three-dimensional position and velocity components. The model predicts the state at the next moment through the state transition equation, and a simplified linear form can be expressed as: , where represents the predicted value of the state vector at time k under the condition of known information at time k - 1; represents the estimated value of the state vector at time k - 1, is the state transition matrix, which describes how the state naturally evolves from time k - 1 to time k, is the control input vector at time k, which can include acceleration instructions or measurement values; is the control input matrix, which maps the control input to the state change. The constructed moving trend prediction model can characterize the typical motion laws of the cluster or a single UAV.

[0041] Collecting the acceleration vector sequences of the cluster node set within a preset time window and inputting them into the moving trend prediction model to generate a predicted moving path; Specifically, when prediction is required, periodically collect from the cluster node set, especially the cluster head node or representative node, their acceleration vector sequences within a recent preset time window , where . These sequence data reflect the recent maneuvering behaviors of the UAVs. Based on this collected historical motion data, use a specific algorithm or technology to train or construct a moving trend prediction model. The moving trend prediction model calculates according to these real-time inputs and its internal state, and outputs the predicted moving path of the relevant UAV or cluster within a future period . The predicted moving path can be represented as a sequence of predicted position points .

[0042] Collect the communication load data of the cluster node set, and generate channel resource pre-allocation parameters in combination with the predicted movement path; Specifically, collect the communication load data from each UAV node or cluster head simultaneously , and the communication load data is a real-time indicator reflecting the data transmission volume and resource occupancy in the current network. Combine and analyze these communication load data with the predicted movement path generated in the previous step. For example, based on the predicted movement path, the future change in the distance between nodes can be estimated, and then the change in link loss can be predicted, or the potential load increase caused by the change in node density in a specific area can be estimated. Considering the current load and the future communication requirements or risks inferred based on the predicted movement path, generate channel resource pre-allocation parameters . The channel resource pre-allocation parameters can be a structured data, including quantitative suggestions or risk assessments for future channel resource requirements, such as the predicted peak bandwidth demand , the link stability risk level .

[0043] Generate a dynamic path planning instruction according to the channel resource pre-allocation parameter and the predicted movement path, and send it to the main cluster head node.

[0044] Specifically, when generating the final dynamic path planning instruction, integrate the predicted movement path and the channel resource pre-allocation parameter . The generated dynamic path planning instruction not only includes basic flight control commands, but also embeds pre-allocation information or suggestions related to communication resource management. This dynamic path planning instruction is then sent to the relevant main cluster head node through the control link.

[0045] Optionally, the trigger link switching instruction includes: Continuously monitor the bit error rate and signal attenuation rate of the main communication channel, and generate an interference level evaluation parameter; Specifically, this step aims to accurately evaluate the current channel condition to determine whether stronger anti-interference measures such as frequency hopping are needed. The UAV node continuously monitors the primary communication channel it is using, focusing on the BER and the signal attenuation rate. The BER directly reflects the accuracy of data transmission. The signal attenuation rate can be indirectly evaluated by comparing the known transmit power Ptx with the real-time measured RSSI, or obtained by analyzing the Channel State Information (CSI). These continuously monitored parameters, possibly combined with other signal quality parameters such as the Signal-to-Noise Ratio (SNR), are fused into a comprehensive interference level evaluation parameter through a preset evaluation function or algorithm. For example, a simple weighted fusion method is as follows: , where, represents the calculated interference level evaluation parameter, and the larger the value, the more serious the interference. The current bit error rate, is the signal attenuation index, is the signal-to-noise ratio, is the corresponding normalization or mapping function that converts the original parameter value into the contribution degree to the interference level, are the corresponding weight coefficients for each item. When this exceeds a preset high interference threshold , it indicates that simply switching to the standby channel may not be sufficient to solve the problem, or the standby channel may also be disturbed, and the frequency hopping mechanism needs to be activated.

[0046] Match the preset frequency hopping sequence library according to the interference level evaluation parameter, select the target frequency hopping mode, generate a link switching instruction and trigger the power compensation instruction for the set of clustered nodes.

[0047] Specifically, when the interference level evaluation parameter exceeds a preset high interference threshold , it indicates that the current channel interference is serious and frequency hopping may need to be activated. At this time, access a preset frequency hopping sequence library that stores various different frequency hopping modes. According to the interference level evaluation parameter and possible other interference characteristic information, select a frequency hopping sequence with the optimal expected anti-interference effect or the most capable of avoiding known interference from the preset frequency hopping sequence library through a matching logic as the target frequency hopping mode. Once the target hopping pattern is selected, corresponding link switching instructions are generated to direct relevant nodes to switch to communicating using the target hopping pattern. Meanwhile, to address potential link instability caused by frequency changes during hopping, when generating the link switching instructions, power compensation instructions for relevant nodes in the cluster node set are triggered simultaneously, such as Figure 3 shown. The power compensation instructions direct the UAV nodes participating in hopping to adjust their transmission power , so as to compensate for the change in channel characteristics at the new frequency, with the aim of enabling the receiving end to maintain a stable received signal power level after the handover.

[0048] Exemplarily, a UAV is in communication. It monitors that the BER of the primary channel sharply rises to 10 -3 , and at the same time, the signal attenuation also significantly increases. The calculated comprehensive interference level evaluation parameter is 0.85, exceeding the threshold for initiating hopping . It is decided to initiate hopping. According to the interference reported by the spectrum sensing module, which mainly concentrates in the frequency band , a target hopping pattern that can effectively avoid the frequency band is selected from the hopping sequence library . While instructing the UAV to initiate hopping, the first hopping frequency fhop1 in is analyzed. Assume that the path loss is expected to be 3 dB higher than the current frequency . Meanwhile, a power compensation instruction is issued, requiring the transmitting UAV to temporarily increase its transmission power by 3 dB. When the UAV hops to , the increased transmission power exactly compensates for the increased path loss, enabling the receiving end to receive a stable signal strength and ensuring smooth communication.

[0049] Optionally, the triggering of the power compensation instruction includes: judging whether the communication frequency band corresponding to the target hopping pattern overlaps with an external interference source, and generating a frequency band conflict identifier according to the judgment result; Specifically, before selecting the target hopping pattern and preparing to trigger the power compensation instruction, an additional check is performed. Using the on-board spectrum sensing module or accessing a known external interference source database, where the external interference source is a wireless signal source outside the UAV cluster that may affect cluster communication, such as other wireless communications, radars, or electronic countermeasure devices. The set of frequency points included in the target hopping pattern is compared with the frequency bands occupied by the known external interference sources . If it is found that Some or all of the frequency points fall within the range of, that is, there is frequency overlap, then a band conflict flag ConflictFlag with a value of "true" is generated. If no overlap is found, ConflictFlag is "false".

[0050] According to the band conflict flag, adjust the power difference between co-frequency nodes in the cluster node set and reduce the bandwidth occupancy ratio of non-critical tasks.

[0051] Specifically, if the band conflict flag ConflictFlag generated in the previous step is "true", this indicates that even if frequency hopping is initiated, it may not be possible to completely avoid external strong interference. In this case, additional measures need to be taken to ensure communication as much as possible. First, it will be identified whether there are other nodes in the cluster node set that are also using frequencies overlapping with the conflict band, in addition to the nodes currently performing frequency hopping and power compensation. These nodes are called co-frequency nodes. If such co-frequency nodes exist, the transmitted power difference between them will be adjusted . The goal of the adjustment is usually to increase the transmitted power of the critical communication link nodes relative to other co-frequency nodes, that is, to improve their signal-to-interference ratio (SIR), so as to gain a relative advantage in the interference environment. For example, non-critical co-frequency nodes can be instructed to reduce their transmitted power, or critical nodes can be instructed to further increase their power within the allowable range. At the same time, in order to reduce the communication volume on the interfered band and reduce the collision probability, the bandwidth occupancy ratio of non-critical tasks will also be temporarily reduced. This means that the channel resources allocated to tasks with lower priorities, such as non-critical tasks like telemetry data upload and log record synchronization, will be reduced, and the precious and possibly interfered bandwidth resources will be preferentially used to transmit high-priority information such as flight control and critical task instructions.

[0052] Optionally, the construction of the interference heat map includes:[[]] Obtain the real-time three-dimensional coordinates of the cluster node set and calculate the signal coverage overlap area between adjacent nodes; Specifically, this step aims to quantify the potential mutual interference between nodes within the cluster. First, it is necessary to obtain the current real-time three-dimensional coordinates of all UAVs in the cluster node set . Based on the coordinates of each UAV node , and its known transmitted power , and the radio wave propagation model, such as considering path loss, shadow fading, etc., the theoretical coverage range of the signal of this node that can reach a certain received power threshold Then, focus on adjacent nodes, i.e., UAV pairs that are relatively close geographically or directly connected in the network topology. For each pair of adjacent nodes, calculate their respective signal coverage ranges. and the intersection region in three-dimensional space, i.e., the signal coverage overlapping region. The size or volume of this overlapping region can be used as an indicator to measure the potential mutual interference intensity between nodes and

[0053] Generate a map of interference hotspots based on the signal coverage overlapping region and external spectrum scanning data; Specifically, this step is to fuse internal interference and external interference information and visualize it. Combine the signal coverage overlapping regions of each pair of adjacent nodes calculated in the previous step or the internal interference risk values derived therefrom with the external spectrum scanning data provided by the airborne spectrum sensing device or the ground station. The external spectrum scanning data provides information about other wireless signal sources present in the operating environment, such as their center frequencies, bandwidths, signal strengths, and possible source directions. Superimpose this internal interference risk information and external interference source information onto a geographical map or network topology map. Generate a map of interference hotspots, i.e., an interference heat map, by setting different colors, grayscale levels, or numerical intensities to represent the total interference levels at different locations or links. This map visually identifies which areas or communication links in the network are suffering from strong interference, forming so-called "interference hotspots".

[0054] Dynamically adjust the topological connection weights of the cluster node set based on the map of interference hotspots.

[0055] Specifically, the generated map of interference hotspots can not only guide power control but also be used to optimize the routing selection of the network. Traditional routing algorithms may mainly select paths based on the hop count or physical distance. In this step, dynamically adjust the weights of the logical topological connections in the network according to the interference distribution shown in the interference heat map. For example, if the heat map shows that the direct communication link between node and node passes through a strong interference area, i.e., the interference value on this link is high, even if and are very close, the weight of this link can be significantly increased. ​When calculating the optimal path, the routing algorithm tends to avoid links with high weights, even if this means choosing a path with more hops but passing through a less interfered area. This adjustment enables data transmission to intelligently bypass severely interfered areas, improving the success rate and reliability of end-to-end communication. Topological connection weight To assign a cost value to the logical connection between nodes in routing calculations, which affects path selection decisions, as Figure 4 shown

[0056] Exemplarily, the interference heat map of a UAV cluster shows that there is a strong interference hotspot in the central area of the cluster, which is jointly caused by an external interference source, such as a ground WiFi hotspot, and the signal overlap of internal nodes. The direct link between nodes U5 and U6 just needs to pass through this hotspot area. Although the physical distance between U5 and U6 is very close, when dynamically adjusting the topological connection weight, the weight of this link U5-U6 is set very high, for example, set to 100. For another detour path, such as U5->U7->U6, although the number of hops is 2, the area it passes through has less interference, and the weights and of these two links may be only 10 respectively. When sending data from U5 to U6, the routing algorithm calculates the path cost and finds that the direct path cost is 100, while the detour path cost is 10 + 10 = 20. Therefore, the algorithm will choose the path U5->U7->U6 to transmit data, thus successfully avoiding the strong interference area.

[0057] Optionally, the dynamically adjusting the topological connection weight of the cluster node set based on the interference hotspot distribution map includes: When detecting that the signal coverage overlap area of cross-cluster nodes exceeds a preset threshold, generating an arbitration request instruction and sending it to the ground control station; Specifically, this step deals with more complex situations that may be caused by inter-cluster interference or resource conflicts. When constructing the interference heat map or performing network topology analysis, special attention is paid to the boundary areas between different clusters. If cross-cluster nodes are detected, that is, UAV nodes belonging to different logical clusters but relatively close geographically (belonging to cluster A) and node (belonging to cluster B), and the severity of the signal coverage overlap area between them exceeds a preset threshold. The degree of overlap can be quantified by an overlap factor , where is the node and The overlap factor between them usually ranges from 0 to 1. The larger the value, the more serious the overlap. and are the effective communication radii of nodes and respectively. is the real-time distance between nodes and . Ensure that the factor is not negative. When the calculated overlap factor exceeds the set threshold , that is, , it indicates that the potential risk of inter-cluster interference is high, or the clustering structure may need to be adjusted. Since this cross-cluster problem involves the interests of multiple clusters, higher-level coordination is required. In this case, the relevant cluster head nodes or the nodes that detect this situation will generate an arbitration request instruction, which is a specific signaling for requesting a decision from a higher-level management unit when local resource conflicts cannot be effectively resolved or a global perspective coordination is needed. This instruction contains conflict descriptions, such as the UAV identifiers involved, the calculated overlap factor , relevant interference measurement values, etc., and is sent to the Ground Control Station (GCS) via the uplink. The GCS is the central unit responsible for monitoring, managing, and making high-level decisions for the entire UAV cluster.

[0058] Receive the global topology optimization instruction fed back by the ground control station and update the communication priority configuration of the clustering node set.

[0059] Specifically, after receiving the arbitration request instruction from the UAV cluster, the ground control station GCS will utilize the global situation information it has and may run a global optimization algorithm to determine the best solution. The goal of this algorithm may be to maximize the overall task utility function of the cluster , while minimizing interference or satisfying the Quality of Service (QoS) constraints. For example, the GCS may need to solve an optimization problem: , where represents the set of priority configurations of all relevant communication tasks or nodes in the cluster. The GCS needs to find the optimal priority configuration. The decision result will be fed back to the relevant UAV nodes or cluster heads in the form of a global topology optimization instruction, which is an adjustment instruction issued by the GCS to improve the network structure, resolve conflicts, or enhance performance as a whole. This instruction usually contains updates to the communication priority configurations of specific nodes or tasks. The priority can be a numerical value, such as an integer or a floating point number. The higher the numerical value, the higher the priority. After receiving the instruction, UAV node k will update its current priority to the new value specified by the GCS This new priority will directly affect the behavior of the node when competing for shared channel resources, such as determining the proportion of bandwidth allocated to it, the size of the backoff window in the channel access mechanism, or its sorting position in the scheduling queue, specifically according to the predefined resource allocation rules to determine. For example, its effective data transmission rate may be related to the priority , Among them, is a function representing the result of resource allocation, is the current competition level of the channel.

[0060] Exemplarily, the boundary nodes of two adjacent UAV clusters and calculate the overlap factor , which exceeds the set threshold . The cluster head of cluster A sends an arbitration request to the GCS. After evaluation, the GCS determines that the task priority of cluster B is higher than that of cluster A. Let B be Set as . To reduce interference, the GCS decides to reduce the communication activity in the conflict area and issue a global topology optimization instruction, clearly sets the task priority for communication in the direction of cluster B to a lower value, for example . After receiving the instruction, according to the preset rules , when its priority is 2, its average backoff time for competing for the channel in this direction will increase significantly, thereby reducing its transmission frequency and potential interference to cluster B, and giving priority to ensuring the communication of the high-priority cluster B.

[0061] Optionally, the pre-allocation parameters of the channel resources include: Calculating the remaining available bandwidth resources according to the pre-allocation parameters of the channel resources; Specifically, this step aims to quantify the tightness of network resources as a basis for optimization decisions. After receiving the dynamic path planning instruction containing the pre-allocation parameters of the channel resources , the main cluster head node will use these parameters to predict the resource requirements in the next period of time. May directly give the predicted bandwidth requirements , or the cluster head needs to estimate according to the predicted topology information and expected service types contained in . At the same time, the cluster head also knows the total capacity of the current channel and the currently allocated or in-use bandwidth Based on this information, the cluster head can calculate the expected remaining available bandwidth resources : , wherein represents the unoccupied transmission capacity remaining on the expected channel after meeting the current and expected future demands. If the calculation result is negative or very small, it indicates that the channel resources may become very tight or insufficient in the future.

[0062] When the remaining available bandwidth resources are lower than the preset threshold, additional channel time slots are preferentially allocated to the nodes corresponding to the power compensation instruction.

[0063] Specifically, a warning threshold for the remaining available bandwidth resources is set . This threshold represents the minimum resource margin that the network can tolerate. The cluster head continuously monitors the calculated . When it is found that is lower than this set threshold , that is it indicates that the network is about to enter or is already in a resource-tight state. In this case, if there are some UAV nodes in the cluster that are executing the power compensation instruction because they have triggered anti-jamming measures, this usually means that these nodes are in a critical communication state or their links are relatively vulnerable. Therefore, in the case of resource tension, the optimized standby channel activation logic or subsequent resource allocation decisions such as allocating additional channel time slots, where the additional channel time slots are additional transmission time segments or frequency resources on the basis of the original allocation, will prioritize meeting the resource requirements of these nodes that are performing power compensation, ensuring that they have sufficient bandwidth to stabilize the link and complete the transmission of critical data, and then considering allocating the remaining resources to other ordinary nodes or activating the standby channel. This preferential allocation mechanism ensures that in the event of network congestion, the most critical or most in-need-of-guarantee communication links can be served preferentially.

[0064] Exemplarily, the total channel capacity managed by a cluster head is 10 Mbps. The currently used bandwidth is 4 Mbps. The received predicts that due to an increase in task requirements in the next 10 seconds, an additional bandwidth is expected. The calculated remaining available bandwidth . Assume that the set remaining bandwidth threshold . Since , the network resources are determined to be tense. At this time, there is exactly one UAV node U5 in the cluster. Because it encounters interference and starts frequency hopping, it is executing a power compensation instruction. According to the optimization logic, when there are any available extra channel time slots, such as obtained by activating a standby channel or other nodes releasing part of the bandwidth, these resources will be preferentially allocated to U5 to ensure the stability of its link after frequency hopping and data transmission, even if other nodes U6 and U7 also have certain bandwidth requests.

[0065] Optionally, adjusting the power difference of co-frequency nodes in the cluster node set further includes: Obtain historical frequency band conflict identifiers, and extract the occurrence frequency and time segment distribution data of frequency band conflict events to obtain a conflict data set; Specifically, this step aims to accumulate empirical data to guide future decisions. During its operation, every occurrence of a frequency band conflict event will be continuously recorded. A frequency band conflict event refers to a situation where the frequency points in the selected target frequency hopping mode overlap with the frequency bands occupied by known external interference sources . For each recorded conflict event, not only the occurrence of the conflict itself is recorded, but also relevant information is associated and recorded to form historical frequency band conflict event data. These data may include the exact timestamp Timestamp of the conflict occurrence, the target frequency hopping mode that caused the conflict , the identified external interference source information, the geographical location Location at the time of the conflict, and the duration Duration of the conflict, etc. By accumulating these data over a long period, it can be analyzed whether certain frequency hopping sequences are more likely to conflict during specific time periods, specific geographical regions, or in the face of specific types of interference sources. For example, the conflict frequency of each frequency hopping sequence at different time periods of a day can be statistically analyzed. .

[0066] Update the priority sorting rules of the frequency hopping sequence library according to the conflict data set.

[0067] Specifically, use the accumulated historical frequency band conflict event data to dynamically optimize the management of the frequency hopping sequence library. Through statistical analysis of historical data, such as calculating the total conflict frequency of each frequency hopping sequence or the conditional conflict probability under specific conditions , the risk of each frequency hopping sequence in actual use can be evaluated. Based on this risk assessment result, the priority sorting rules of each sequence in the frequency hopping sequence library will be dynamically updated. This rule determines which sequences will be preferentially considered when a target frequency hopping mode needs to be selected. For example, frequency hopping sequences with high historical conflict frequencies or high predicted conflict probabilities under current conditions can be Priority is lowered, while the priority of sequences with good performance and few conflict records is raised. When a frequency hopping sequence needs to be selected, it will be preferentially selected from sequences with higher priority so as to more likely avoid known and frequent frequency band conflicts and improve the success rate of frequency hopping anti-interference.

[0068] Exemplarily, after running for one month, the statistical data of historical frequency band conflict event shows that the frequency hopping sequence has a significantly higher frequency of conflict with a certain specific communication signal frequency band between 2 pm and 4 pm every day, especially over urban areas, than other sequences. According to this statistical result, the priority sorting rule of the frequency hopping sequence library is updated. Assuming the priority range is 0 - 1, the priority during the period from 2 pm to 4 pm is significantly reduced from the original 0.8 to 0.2. When a UAV cluster over a city needs to start frequency hopping at 3 pm on a certain day later, when matching the frequency hopping sequence library, it will tend to select other sequences with higher priority because the priority under this condition is extremely low, such as or , even if the performance indicators in other aspects may be good.

[0069] Based on the same inventive concept, the present invention also provides a UAV cluster interaction method and system based on an ad hoc network. The system includes: An information acquisition and clustering module, configured to acquire the positioning information, remaining energy parameters, and real-time task types of the UAV group, and generate a clustering node set; A movement prediction and planning module, configured to predict the movement direction of the cluster based on the motion acceleration data of the clustering node set, and generate a dynamic path planning instruction; A channel allocation module, configured to allocate a primary communication channel according to the dynamic path planning instruction, activate a standby channel, and generate a channel allocation result; A link monitoring and switching module, configured to monitor the signal quality parameters of the primary communication channel in the channel allocation result, and trigger a link switching instruction when interference is detected; A topology and power control module, configured to construct an interference heat map according to the topological relationship of the clustering node set, and generate a power adjustment instruction.

Claims

1. A drone cluster interaction method based on a self-organizing network, characterized in that: The method comprises: Obtain the positioning information, remaining energy parameters and real-time task types of the drone group, and generate a cluster node set; Extracting motion acceleration data of the drone cluster based on the cluster node set, predicting the cluster movement direction, and generating dynamic path planning instructions; Allocate a primary communication channel according to the dynamic path planning instruction and generate a channel allocation result; monitoring the signal quality parameters of the primary communication channel according to the channel allocation result, and triggering a link switching instruction when interference is detected; An interference heat map is constructed according to the topological relationship of the clustered node set, and a power adjustment instruction is generated.

2. The method for drone cluster interaction based on a self-organizing network as claimed in claim 1, characterized in that: Generating a clustering node set includes: Extracting the correlation parameter between the positioning information and the real-time task type, and screening out a set of candidate nodes that meet the preset cluster head conditions; Determining a primary cluster head node from the candidate node set based on the residual energy parameter and a preset mobility stability coefficient; According to the communication radius of the main cluster head node and a preset signal strength threshold, the member range of the sub-cluster node set is divided to generate a sub-cluster node set.

3. The method for drone cluster interaction based on a self-organizing network as claimed in claim 2, characterized in that: The generating of dynamic path planning instructions comprises: Collect historical movement data of drone clusters and build a movement trend prediction model; Collecting the acceleration vector sequence of the clustered node set within a preset time window and inputting it into the movement trend prediction model to generate a predicted movement path; Collecting communication load data of the clustered node set and generating channel resource pre-allocation parameters in combination with the predicted moving path; A dynamic path planning instruction is generated according to the channel resource pre-allocation parameters and the predicted moving path, and is sent to the main cluster head node.

4. The method for drone cluster interaction based on a self-organizing network as claimed in claim 1, characterized in that: The trigger link switching instruction includes: Continuously monitoring the bit error rate and signal attenuation rate of the main communication channel to generate interference level assessment parameters; A preset frequency hopping sequence library is matched according to the interference level assessment parameter, a target frequency hopping mode is selected, a link switching instruction is generated, and a power compensation instruction of the clustered node set is triggered.

5. The method for drone cluster interaction based on a self-organizing network as claimed in claim 4, characterized in that: The trigger power compensation instruction includes: Determine whether the communication frequency band corresponding to the target frequency hopping mode overlaps with the external interference source, and generate a frequency band conflict identifier according to the determination result; According to the frequency band conflict identifier, the power difference of the same-frequency nodes of the clustered node set is adjusted and the proportion of non-critical task bandwidth is reduced.

6. The method for drone cluster interaction based on a self-organizing network as claimed in claim 1, characterized in that: The construction of the interference heat map includes: Obtaining the real-time three-dimensional coordinates of the clustered node set, and calculating the signal coverage overlap area between adjacent nodes; Generate an interference hotspot distribution map according to the signal coverage overlap area and external spectrum scanning data; The topological connection weight of the clustering node set is dynamically adjusted based on the interference hotspot distribution map.

7. The method for drone cluster interaction based on a self-organizing network as claimed in claim 6, characterized in that: The dynamically adjusting the topological connection weight of the clustered node set based on the interference hotspot distribution map includes: When it is detected that the signal coverage overlap area across cluster nodes exceeds a preset threshold, an arbitration request instruction is generated and sent to the ground control station; Receive the global topology optimization instruction fed back by the ground control station, and update the communication priority configuration of the cluster node set.

8. The method for drone cluster interaction based on a self-organizing network as claimed in claim 3, characterized in that: The pre-allocation of the channel resource parameters comprises: Calculate the remaining available bandwidth resources according to the channel resource pre-allocation parameters; When the remaining available bandwidth resources are lower than a preset threshold, additional channel time slots are preferentially allocated to the node corresponding to the power compensation instruction.

9. The method for drone cluster interaction based on a self-organizing network as claimed in claim 5, characterized in that: The adjusting the power difference of the same-frequency nodes of the clustered node set further includes: Obtain historical frequency band conflict identifiers, and extract the frequency and time distribution data of frequency band conflict events to obtain a conflict data set; According to the conflicting data set, a priority sorting rule of the frequency hopping sequence library is updated.

10. A drone cluster interaction system based on a self-organizing network, applied to execute the drone cluster interaction method based on a self-organizing network as claimed in any one of claims 1 to 9, characterized in that: The system comprises: The information acquisition and clustering module is used to obtain the positioning information, remaining energy parameters and real-time task types of the drone group, and generate a clustering node set; A movement prediction and planning module, used to predict the movement direction of the cluster based on the movement acceleration data of the cluster node set and generate dynamic path planning instructions; A channel allocation module, used to allocate a primary communication channel and activate a backup channel according to the dynamic path planning instruction, and generate a channel allocation result; A link monitoring and switching module, configured to monitor the signal quality parameters of the primary communication channel in the channel allocation result and trigger a link switching instruction when interference is detected; The topology and power control module is used to construct an interference heat map according to the topological relationship of the clustered node set and generate a power adjustment instruction.

Images