Unmanned aerial vehicle flight communication state monitoring method and system based on multiple links

Abstract

The invention discloses an unmanned aerial vehicle flight communication state monitoring method and system based on multiple links, and relates to the field of unmanned aerial vehicle communication, and the method comprises the steps: collecting physical layer parameters of at least two heterogeneous communication links at the same time; for each heterogeneous communication link, mapping physical layer parameters into a virtual vibration displacement sequence; extracting the amplitude of the virtual vibration displacement sequence, and performing frequency domain transformation on the virtual vibration displacement sequence of the at least two heterogeneous communication links to obtain a corresponding frequency spectrum function; calculating the cross-power spectral density between the spectrum functions of different heterogeneous communication links, and determining a coherence function and the maximum coherence value of the coherence function; acquiring flight attitude data of the unmanned aerial vehicle, and dynamically adjusting a preset homologous interference judgment threshold value; and generating a communication state monitoring conclusion based on the maximum coherent value, the adjusted homologous interference judgment threshold and the amplitude, and executing a corresponding link control strategy. The flight safety of the unmanned aerial vehicle can be effectively guaranteed.

Description

A method and system for monitoring the flight communication status of unmanned aerial vehicles (UAVs) based on multi-link communication. Technical Field

[0001] This application relates to the field of unmanned aerial vehicle (UAV) communication, and more particularly to a method and system for monitoring the flight communication status of UAVs based on multiple links. Background Technology

[0002] With the rapid development of UAV technology and the continuous expansion of its application scenarios, UAV communication systems have become a key element in ensuring flight safety and mission execution. To improve communication reliability, modern UAVs generally adopt a multi-link communication architecture, configuring multiple heterogeneous communication links (such as 5G links, 2.4G radio links, satellite links, etc.) as redundancy backups. In traditional communication system design concepts, these communication links of different frequency bands and standards are considered independent of each other, with no correlation between them. Based on this assumption, a multi-link system can quickly switch to a backup link when a single link fails, thereby ensuring the continuity and stability of communication.

[0003] However, in practical applications, especially in complex urban electromagnetic environments, extreme weather conditions, or special geographical environments, multiple seemingly independent heterogeneous communication links are often affected by the same external interference source. For example, external factors such as broadband frequency sweep interference, electromagnetic storms, and strong lightning interference may act on multiple links simultaneously, causing the communication quality of each link to decline almost synchronously. However, in the initial stage, the indicators of a single link may still be within the normal range or only show slight fluctuations, making it difficult to trigger traditional alarm mechanisms.

[0004] Existing UAV communication status monitoring methods primarily employ a single-link independent monitoring approach. This involves monitoring the physical layer parameters (such as signal-to-noise ratio, bit error rate, and packet loss rate) of each link separately and comparing them to preset thresholds. When the parameters of a link exceed the threshold, the system triggers link switching or an alarm. However, this method cannot identify multi-link coordinated degradation trends caused by interference from the same source. By the time the monitoring system finally detects that all links have deteriorated severely simultaneously, the optimal response time has often been missed, leading to communication system collapse and seriously threatening the flight safety of the UAV.

[0005] Therefore, existing monitoring methods cannot identify the common source interference of the external environment on multiple links in the early stages when communication indicators appear normal or fluctuate slightly. As a result, they cannot predict and actively avoid the risk of communication interruption in advance, which seriously affects the flight safety of UAVs. Summary of the Invention

[0006] This application provides a method and system for monitoring the flight communication status of unmanned aerial vehicles (UAVs) based on multiple links. It is used to identify the common source interference of the external environment on multiple links in the early stage when the communication indicators seem normal or fluctuate slightly, so as to realize the early prediction and active avoidance of communication interruption risk and effectively ensure the flight safety of UAVs.

[0007] To achieve the above objectives, the embodiments of this application adopt the following technical solutions: Firstly, a method for monitoring the flight communication status of a UAV based on multiple links is provided, applied to a UAV, wherein the UAV is configured with at least two heterogeneous communication links. The method includes: simultaneously acquiring physical layer parameters of at least two heterogeneous communication links; mapping the physical layer parameters to a virtual vibration displacement sequence for each heterogeneous communication link; extracting the amplitude of the virtual vibration displacement sequence within a preset time window, and performing a frequency domain transformation on the virtual vibration displacement sequences of the at least two heterogeneous communication links to obtain the corresponding spectral functions; calculating the cross-power spectral density between the spectral functions of different heterogeneous communication links, and determining the coherence function and the maximum coherence value of the coherence function based on the cross-power spectral density; acquiring the flight attitude data of the UAV, and dynamically adjusting a preset co-source interference judgment threshold based on the flight attitude data; generating a communication status monitoring conclusion based on the maximum coherence value, the adjusted co-source interference judgment threshold, and the amplitude; and executing a corresponding link control strategy based on the communication status monitoring conclusion, wherein the link control strategy includes a position avoidance strategy, a link switching strategy, and a maintenance strategy.

[0008] In one possible implementation of the first aspect, the physical layer parameters include real-time signal-to-noise ratio, transmission delay, and available bandwidth. For each heterogeneous communication link, the physical layer parameters are mapped to a virtual vibration displacement sequence, including: determining virtual mass parameters for each heterogeneous communication link; mapping the available bandwidth of the heterogeneous communication link to a stiffness coefficient and the transmission delay to a damping coefficient, wherein the stiffness coefficient is positively correlated with the available bandwidth and the damping coefficient is positively correlated with the transmission delay; using the reciprocal of the real-time signal-to-noise ratio of the heterogeneous communication link as the excitation force function; constructing a second-order differential equation, which includes the virtual mass parameter, stiffness coefficient, damping coefficient, and excitation force function; and iteratively solving the second-order differential equation to generate the virtual vibration displacement sequence of the heterogeneous communication link.

[0009] In another possible implementation of the first aspect, the heterogeneous communication links include a first heterogeneous communication link and a second heterogeneous communication link. The cross-power spectral density between the spectral functions of the different heterogeneous communication links is calculated, and the coherence function and its maximum coherence value are determined based on the cross-power spectral density. This includes: performing windowed Fourier transforms on the virtual vibration displacement sequences of the first and second heterogeneous communication links respectively to obtain a first spectral function and a second spectral function; calculating the cross-power spectral density based on the first and second spectral functions; calculating the first auto-power spectral density of the first spectral function and the second auto-power spectral density of the second spectral function respectively; calculating the coherence function based on the cross-power spectral density, the first auto-power spectral density, and the second auto-power spectral density; and iterating through the coherence values ​​of the coherence function at different frequency points to extract the maximum coherence value.

[0010] In another possible implementation of the first aspect, the coherence function is calculated based on the cross-power spectral density, the first self-power spectral density, and the second self-power spectral density, including: calculating the square of the modulus of the cross-power spectral density; calculating the product of the first self-power spectral density and the second self-power spectral density; and dividing the square of the modulus of the cross-power spectral density by the product to obtain the coherence function.

[0011] In another possible implementation of the first aspect, the flight attitude data includes three-axis acceleration and three-axis angular velocity. The preset threshold for determining co-source interference is dynamically adjusted based on the flight attitude data, including: calculating the total acceleration of the UAV based on the three-axis acceleration; calculating the maneuver intensity factor based on the total acceleration; calculating the angular velocity magnitude of the UAV based on the three-axis angular velocity; calculating the attitude change rate based on the angular velocity magnitude; calculating the comprehensive maneuver index based on the maneuver intensity factor and the attitude change rate; calculating the threshold adjustment coefficient based on the comprehensive maneuver index, wherein a higher comprehensive maneuver index results in a larger threshold adjustment coefficient; and multiplying the preset threshold for determining co-source interference by the threshold adjustment coefficient to obtain the adjusted threshold for determining co-source interference.

[0012] In another possible implementation of the first aspect, a communication status monitoring conclusion is generated based on the maximum coherence value, the adjusted same-source interference judgment threshold, and the amplitude. This includes: determining the communication status monitoring conclusion as regional electromagnetic environment deterioration when the maximum coherence value exceeds the adjusted same-source interference judgment threshold and the amplitudes of the virtual vibration displacement sequences of at least two heterogeneous communication links all exceed a preset amplitude threshold; determining the communication status monitoring conclusion as a single-link failure when the maximum coherence value does not exceed the same-source interference judgment threshold and the amplitude of the virtual vibration displacement sequence of a single heterogeneous communication link exceeds a preset amplitude threshold; and determining the communication status monitoring conclusion as normal communication when the maximum coherence value does not exceed the same-source interference judgment threshold and the amplitudes of the virtual vibration displacement sequences of all heterogeneous communication links do not exceed a preset amplitude threshold.

[0013] In another possible implementation of the first aspect, based on the communication status monitoring conclusion, a corresponding link control strategy is executed, including: if the communication status monitoring conclusion is that the regional electromagnetic environment has deteriorated, the link switching strategy is prohibited and a location avoidance strategy is executed; if the communication status monitoring conclusion is that a single link has failed, the link switching strategy is executed to switch the communication service to a healthy link; if the communication status monitoring conclusion is that the communication status is normal, a maintenance strategy is executed to maintain the current communication configuration.

[0014] In another possible implementation of the first aspect, the position avoidance strategy includes: controlling the UAV to perform a vertical climb operation; continuously monitoring the maximum coherence value and the amplitude of the virtual vibration displacement sequence during the climb; and stopping the climb operation when the maximum coherence value drops below the threshold for determining common source interference and the amplitude of the virtual vibration displacement sequence drops below a preset amplitude threshold.

[0015] Secondly, this application provides a drone, including: a memory configured to store instructions; and an onboard processor configured to retrieve the instructions from the memory and, when executing the instructions, to implement the aforementioned multi-link-based drone flight communication status monitoring method.

[0016] Thirdly, this application provides a multi-link-based UAV flight communication status monitoring system, comprising: a UAV equipped with at least two heterogeneous communication links; and a ground control station that communicates with the UAV through at least two heterogeneous communication links.

[0017] The above technical solution maps the physical layer parameters of heterogeneous communication links to virtual vibration displacement sequences. By utilizing frequency domain analysis and coherence function calculation, a quantitative assessment of the correlation between multiple links is achieved, effectively solving the problem of existing technologies being unable to identify interference from the same source. Specifically, physical layer parameters of at least two heterogeneous communication links are collected simultaneously, ensuring data time synchronization. By mapping physical layer parameters to virtual vibration displacement sequences, communication quality indicators are transformed into physical quantities suitable for vibration analysis, allowing for in-depth analysis based on mature signal processing theories. After extracting amplitude information within a preset time window, the virtual vibration displacement sequence is frequency-domain transformed to obtain a spectral function. Converting the time-domain signal to the frequency domain reveals the frequency characteristics of interference on different links more clearly. Calculating the cross-power spectral density between the spectral functions of different heterogeneous communication links and determining the coherence function and maximum coherence value accordingly allows for precise quantification of the correlation between different links. When multiple links are affected by the same interference source, their coherence values ​​will significantly increase. Even if the indicators of a single link have not yet triggered the traditional alarm threshold, this solution can identify the existence of interference from the same source in advance through abnormal changes in coherence values. Furthermore, considering that changes in UAV flight attitude affect antenna orientation and signal propagation characteristics, flight attitude data is introduced to dynamically adjust the threshold for detecting co-source interference, avoiding misjudgments caused by flight maneuvers and improving the accuracy and adaptability of monitoring. Communication status monitoring conclusions can issue early warnings at the early stages of coordinated degradation in multiple links, gaining valuable response time. Executing corresponding link control strategies based on monitoring conclusions achieves proactive prevention, effectively avoiding communication avalanche phenomena, significantly improving the reliability and security of the UAV multi-link communication system, and thus effectively ensuring the flight safety of the UAV.

[0018] Other features and advantages of the embodiments of this application will be described in detail in the following detailed description section. Attached Figure Description

[0019] Figure 1 is a flowchart illustrating a method for monitoring the flight communication status of a UAV based on multiple links according to an embodiment of this application; Figure 2 is a flowchart illustrating a method for determining the climb height for a vertical climb operation according to an embodiment of this application; Figure 3 is a structural diagram illustrating a system for monitoring the flight communication status of a UAV based on multiple links according to an embodiment of this application. Detailed Implementation

[0020] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are only for illustration and explanation of the embodiments of this application and are not intended to limit the embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0021] It should be noted that if the embodiments of this application involve directional indicators (such as up, down, left, right, front, back, etc.), the directional indicators are only used to explain the relative positional relationship and movement of each component in a certain specific posture (as shown in the figure). If the specific posture changes, the directional indicators will also change accordingly.

[0022] Furthermore, if the embodiments of this application involve descriptions such as "first" or "second," these descriptions are for descriptive purposes only and should not be construed as indicating or implying their relative importance or implicitly specifying the number of technical features indicated. Therefore, features defined with "first" or "second" may explicitly or implicitly include at least one of those features. Additionally, the technical solutions of various embodiments can be combined with each other, but this must be based on the ability of those skilled in the art to implement them. If the combination of technical solutions is contradictory or impossible to implement, it should be considered that such a combination of technical solutions does not exist and is not within the scope of protection claimed in this application.

[0023] Figure 1 schematically illustrates a flowchart of a multi-link-based UAV flight communication status monitoring method according to an embodiment of this application. As shown in Figure 1, this application provides a multi-link-based UAV flight communication status monitoring method, applied to a UAV, which is configured with at least two heterogeneous communication links. The method may include the following steps.

[0024] S110. Simultaneously acquire physical layer parameters of at least two heterogeneous communication links; S120. For each heterogeneous communication link, map the physical layer parameters into a virtual vibration displacement sequence; S130. Extract the amplitude of the virtual vibration displacement sequence within a preset time window, and perform frequency domain transformation on the virtual vibration displacement sequences of at least two heterogeneous communication links to obtain the corresponding spectrum function; S140. Calculate the cross power spectral density between the spectrum functions of different heterogeneous communication links, and determine the coherence function and the maximum coherence value of the coherence function based on the cross power spectral density; S150. Acquire the flight attitude data of the UAV, and dynamically adjust the preset co-source interference judgment threshold based on the flight attitude data; S160. Generate a communication status monitoring conclusion based on the maximum coherence value, the adjusted co-source interference judgment threshold, and the amplitude; S170. Execute the corresponding link control strategy based on the communication status monitoring conclusion, wherein the link control strategy includes a position avoidance strategy, a link switching strategy, and a maintenance strategy.

[0025] In this embodiment, at least two heterogeneous communication links include at least two of the following: cellular network communication links, private band data radio communication links, and satellite communication links.

[0026] Specifically, the UAV flight communication status monitoring method based on multiple links provided in this embodiment is applied to UAV systems configured with at least two heterogeneous communication links. These heterogeneous communication links can include communication methods of different standards and frequency bands, such as 5G cellular network links, 2.4GHz data radio links, and satellite communication links. The UAV's onboard processor acquires communication status data in real time through the communication modules of each link and executes the various steps of this method.

[0027] In practical applications, UAVs synchronously collect physical layer parameters through the hardware modules of each communication link. These physical layer parameters include three core indicators: real-time signal-to-noise ratio (SNR), transmission delay, and available bandwidth. Real-time SNR reflects the ratio of the current received signal quality to the noise level, typically measured in decibels (dB), and is directly measured by the radio frequency (RF) front-end of the communication module. Transmission delay refers to the round-trip time from sending a data packet to receiving an acknowledgment; it is obtained by embedding a timestamp in the data packet and calculating the round-trip delay, and is measured in milliseconds. Available bandwidth represents the data transmission rate currently supported by the link, obtained through the rate negotiation mechanism of the link layer protocol, and is measured in Mbps.

[0028] To ensure data time synchronization, a unified sampling clock is used to trigger parameter acquisition across all links, with a sampling frequency set to 10Hz, meaning that all three parameters of all links are synchronously acquired every 100 milliseconds. For example, at a certain moment, the 5G link measures a signal-to-noise ratio of 25dB, a transmission delay of 30ms, and an available bandwidth of 50Mbps; the 2.4G radio link measures a signal-to-noise ratio of 18dB, a transmission delay of 45ms, and an available bandwidth of 10Mbps; and the satellite link measures a signal-to-noise ratio of 12dB, a transmission delay of 200ms, and an available bandwidth of 2Mbps. The acquired data is stored in a circular buffer in time-series format, retaining the most recent 30 seconds of historical data for subsequent analysis. This synchronous acquisition mechanism ensures strict alignment of data from different links in the time dimension, laying the foundation for subsequent correlation analysis and avoiding coherence calculation errors caused by time deviations.

[0029] For each heterogeneous communication link, the acquired physical layer parameters are mapped to a virtual vibration displacement sequence. This mapping process draws inspiration from the mathematical model of a mechanical vibration system, analogizing the communication link to a forced vibration system. First, virtual mass parameters are set for each link. This parameter is a normalization constant, typically set to 1 kg. Next, the available bandwidth of the link is mapped to a stiffness coefficient. The mapping relationship is ,in For available bandwidth, The proportionality coefficient has a value of 0.1N / (m·Mbps). The stiffness coefficient reflects the link's resilience; the larger the bandwidth, the stronger the system's ability to resist interference.

[0030] Mapping transmission delay to damping coefficient The mapping relationship is ,in For transmission delay, The proportionality coefficient is 0.05 N·s / (m·ms). The damping coefficient characterizes the energy dissipation properties of the link; the greater the delay, the slower the system response. The reciprocal of the real-time signal-to-noise ratio is used as the excitation force function. ,Right now ,in For a moment The signal-to-noise ratio, This is the force scaling factor, with a value of 100 N. The lower the signal-to-noise ratio, the greater the excitation force, indicating stronger interference to the link.

[0031] Based on the above mapping, construct the second-order differential equation: in The virtual vibration displacement is calculated using the fourth-order Runge-Kutta method. The differential equation is solved numerically with a time step of 0.1 seconds, and the displacement value at each sampling moment is obtained iteratively, forming a virtual vibration displacement sequence. For example, for a 5G link, at a certain moment with a bandwidth of 50Mbps, a delay of 30ms, and a signal-to-noise ratio of 25dB, the calculated stiffness coefficient is 5N / m, the damping coefficient is 1.5N·s / m, and the excitation force is 4N. Through iterative solution, the virtual displacement at that moment is obtained as 0.8m. This mapping method transforms abstract communication parameters into intuitive physical quantities, enabling in-depth analysis using mature vibration signal analysis theory and providing a unified mathematical framework for subsequent frequency domain analysis and coherence calculations.

[0032] The amplitude of the virtual vibration displacement sequence within a preset time window is extracted, and the sequence is then subjected to frequency domain transformation. The preset time window length is set to 10 seconds, containing 100 sampling points. For the virtual vibration displacement sequence of each link, the maximum amplitude within that time window is first calculated, i.e., the maximum absolute value of the displacement in the sequence. This amplitude reflects the maximum degree of interference experienced by the link during that period. For example, the virtual displacement sequence of a 5G link within a 10-second window is as follows: Then the amplitude is 1.2m.

[0033] Next, a frequency domain transformation is performed on the virtual vibration displacement sequence. First, the Hanning window function is applied to window the sequence to reduce spectral leakage. The Hanning window function is... ,in The sampling point number, Where is the window length. The sequence after windowing is... Then, a Fast Fourier Transform (FFT) is performed on the windowed sequence to obtain the spectral function. ,in The frequency variable is defined with a frequency resolution of 0.1 Hz. The spectral function is in complex form and includes amplitude and phase information. The above operation is performed on at least two heterogeneous communication links to obtain their respective spectral functions. For example, the spectral function of a 5G link is... The spectrum function of the 2.4G link is .

[0034] Frequency domain transformation converts time-domain signals to the frequency domain, clearly revealing the energy distribution of different frequency components and providing a foundation for identifying the frequency characteristics of interference from the same source. When multiple links are affected by the same interference source, their spectral functions will exhibit similar peak characteristics within a specific frequency range. This similarity serves as the basis for subsequent coherence function calculations.

[0035] The cross-power spectral density between the spectral functions of different heterogeneous communication links is calculated, and the coherence function and maximum coherence value are determined accordingly. Taking the first heterogeneous communication link (e.g., a 5G link) and the second heterogeneous communication link (e.g., a 2.4G link) as examples, their spectral functions have been obtained. and Cross-power spectral density is defined as the frequency domain representation of the cross-correlation of two spectral functions, and its calculation formula is: ,in express The conjugate complex number of the two links. The cross-power spectral density reflects the correlation and phase relationship between the two links at various frequency points.

[0036] Next, the self-power spectral density of each link is calculated. The self-power spectral density of the first link is: The self-power spectral density of the second link is The self-power spectral density represents the energy distribution of a single link at various frequency points. Based on the cross-power spectral density and the self-power spectral density, the coherence function is calculated. The steps for calculating the coherence function are as follows: First, calculate the square of the modulus of the cross-power spectral density. Then calculate the product of the self-power spectral densities of the two links. Finally, dividing the former by the latter yields the coherence function: The coherence function ranges from 0 to 1, representing the frequency differences between two links. The degree of linear correlation at each frequency. When the coherence value is close to 1, it indicates that the two links are highly correlated at that frequency and may be affected by the same interference source; when the coherence value is close to 0, it indicates that the two links are independent at that frequency. The maximum coherence value is extracted by iterating through the values ​​of the coherence function at all frequency points. For example, at a frequency of 0.5Hz, the calculated coherence value is 0.85, which is the maximum value among all frequency points; therefore, the maximum coherence value is 0.85. As a comprehensive indicator of multi-link correlation, the maximum coherence value can keenly detect link-coordinated degradation caused by interference from the same source. Even if the indicator of a single link has not yet triggered a traditional alarm, this indicator can still issue an early warning.

[0037] The system acquires the UAV's flight attitude data and dynamically adjusts the preset threshold for detecting interference from the same source based on this data. The flight attitude data includes three-axis acceleration and three-axis angular velocity, which are obtained in real-time by the onboard inertial measurement unit (IMU). The three-axis acceleration includes... , , The unit is m / s²; triaxial angular velocity includes , , The unit is rad / s.

[0038] First, calculate the total acceleration of the drone's body based on the three-axis acceleration. Then, the computer's dynamic intensity factor. ,in The gravitational acceleration is 9.8 m / s², and this factor characterizes the intensity of the UAV's maneuvers. Next, the angular velocity modulus is calculated based on the three-axis angular velocities. And calculate the rate of change of attitude. ,in The reference angular velocity is 0.1 rad / s.

[0039] Calculate the comprehensive maneuver index based on the maneuver intensity factor and attitude change rate. The weighting coefficients reflect the relative importance of acceleration and angular velocity on communication. A higher comprehensive maneuverability index indicates more violent UAV flight maneuvers, greater changes in antenna pointing and signal propagation paths, potentially leading to synchronized fluctuations in communication quality across multiple links. However, these fluctuations are not caused by external interference from the same source. To avoid misjudgment, the threshold for determining interference from the same source needs to be dynamically adjusted based on the comprehensive maneuverability index.

[0040] Calculate the threshold adjustment coefficient When the comprehensive maneuverability index is 0, the adjustment coefficient is 1, and the threshold remains unchanged; when the comprehensive maneuverability index increases, the adjustment coefficient increases, and the threshold increases accordingly. The preset threshold for determining co-source interference is set. (A value typically taken as 0.7) is multiplied by an adjustment factor to obtain the adjusted threshold. For example, when the drone is in stable flight, the comprehensive maneuverability index is 0.2, the adjustment coefficient is 1.1, and the adjusted threshold is 0.77; when the drone performs a sharp turn, the comprehensive maneuverability index is 1.5, the adjustment coefficient is 1.75, and the adjusted threshold is 1.225 (but the maximum coherence value is 1, so the actual threshold is 1). The dynamic threshold adjustment mechanism effectively distinguishes between normal fluctuations caused by flight maneuvers and abnormal correlations caused by external interference, significantly improving the accuracy and robustness of the monitoring system and reducing the false alarm rate.

[0041] Based on the maximum coherence value, the adjusted same-source interference threshold, and the amplitude, a communication status monitoring conclusion is generated. The monitoring conclusions are divided into three categories: deterioration of the regional electromagnetic environment, single-link failure, and normal communication status. The judgment logic is as follows: First, determine whether the maximum coherence value exceeds the adjusted same-source interference threshold. If it does, further check whether the amplitudes of the virtual vibration displacement sequences of at least two heterogeneous communication links all exceed a preset amplitude threshold (usually set to 1.0m). If both conditions are met simultaneously, i.e. And the amplitude of at least two links is If the amplitude is 1.3m, it is determined to be a deterioration of the regional electromagnetic environment. This indicates that multiple links are affected by the same external interference source, resulting in a coordinated decline in communication quality, and the degree of decline has reached a significant level. For example, when the maximum coherence value is 0.88, the adjusted threshold is 0.77, the amplitude of the 5G link is 1.3m, the amplitude of the 2.4G link is 1.5m, and the amplitude of the satellite link is 1.1m, then it is determined to be a deterioration of the regional electromagnetic environment.

[0042] If the maximum coherence value does not exceed the threshold, i.e. This indicates that each link is relatively independent, and the amplitude of a single link is checked. If the amplitude of a certain link exceeds a preset amplitude threshold, while the amplitudes of other links are normal, it is determined to be a single-link failure. This indicates that there is a problem with the link itself, such as hardware failure or local interference, rather than interference from the same source. For example, if the maximum coherence value is 0.65, the amplitude of the 5G link is 1.4m, and the amplitudes of other links are all less than 1.0m, then it is determined to be a single-link failure of the 5G link. If the maximum coherence value does not exceed the threshold, and the amplitude of all links does not exceed the preset amplitude threshold, then the communication status is determined to be normal. For example, if the maximum coherence value is 0.60, and the amplitude of all links is less than 0.8m, then the communication status is determined to be normal. This hierarchical judgment mechanism can accurately distinguish different types of communication anomalies, providing accurate decision-making basis for subsequent targeted control strategies, avoiding a single processing method, and improving the system's intelligence level.

[0043] Based on the communication status monitoring results, corresponding link control strategies are implemented. These strategies include three categories: location avoidance strategies, link switching strategies, and maintenance strategies. When the communication status monitoring results indicate a deterioration in the regional electromagnetic environment, link switching strategies are prohibited because all links are affected by interference from the same source. Switching to a backup link will not improve communication quality and may even lead to communication interruptions due to frequent switching.

[0044] At this point, a position avoidance strategy is implemented, with the following steps: The UAV is controlled to perform a vertical climb operation, with a climb rate set to 2 m / s. The principle of the climb is to change the UAV's spatial position, moving away from the interference source or changing the signal propagation path, thereby reducing the interference intensity. In one embodiment of this example, during the climb, the maximum coherence value and the amplitude of the virtual vibration displacement sequence are continuously monitored, with the sampling frequency maintained at 10 Hz. When the maximum coherence value is detected to decrease below the adjusted threshold for determining the same source interference, and the amplitude of the virtual vibration displacement sequence of all links decreases below the preset amplitude threshold, it is determined that the UAV has escaped the interference area, the climb operation is immediately stopped, and an alarm message is sent to the ground control station, recording the time, location, and duration of the interference. For example, if the UAV detects a deterioration in the regional electromagnetic environment at an altitude of 100 m and begins to climb, and at 150 m, the maximum coherence value drops to 0.68, and the amplitude of all links drops below 0.9 m, meeting the exit conditions, the UAV stops climbing and hovers.

[0045] When communication status monitoring indicates a single-link failure, a link switching strategy is executed. Communication services are automatically switched from the faulty link to a healthy link. The switching process employs a "build first, then disconnect" mechanism to ensure communication continuity. For example, if a 5G link failure is detected, services are switched to a 2.4G radio link, with switching latency controlled within 200ms. After the switch is complete, the faulty link undergoes a self-check and restart attempt. If it recovers, it is reinstated into the available link pool.

[0046] When the communication status monitoring concludes that the communication status is normal, a maintenance strategy is executed to keep the current communication configuration unchanged and continue monitoring the status of each link. This maintenance strategy avoids unnecessary link switching and flight maneuvers, reducing system energy consumption and mechanical wear. Through the above-mentioned classification control strategy, the system can take optimal countermeasures according to different communication anomalies, avoiding ineffective switching under co-source interference and ensuring rapid recovery in the event of a single link failure, significantly improving the reliability and intelligence level of the UAV multi-link communication system.

[0047] This embodiment maps the physical layer parameters of heterogeneous communication links to virtual vibration displacement sequences and utilizes frequency domain analysis and coherence function calculation to achieve quantitative assessment of the correlation between multiple links, effectively solving the problem that existing technologies cannot identify interference from the same source. The synchronous acquisition mechanism ensures data temporal consistency, the virtual vibration model transforms communication parameters into analyzable physical quantities, frequency domain transformation reveals the frequency characteristics of interference, and cross-power spectral density and coherence function accurately quantify the correlation between links. When multiple links are affected by the same interference source, even if a single link's index has not yet triggered a traditional alarm, an abnormal increase in the maximum coherence value can identify the existence of interference from the same source in advance, buying valuable time for response. The dynamic threshold adjustment mechanism, combined with flight attitude data, effectively distinguishes between normal fluctuations caused by flight maneuvers and abnormal correlations caused by external interference, significantly reducing the false alarm rate. The hierarchical monitoring conclusions and corresponding control strategies enable precise policy implementation: the position avoidance strategy uses spatial maneuvers to escape the interference area, the link switching strategy ensures rapid recovery in the event of a single link failure, and the maintenance strategy avoids unnecessary system disturbances. The overall solution can issue early warnings in the early stages when communication indicators appear normal or fluctuate slightly, proactively preventing communication avalanche phenomena, significantly improving the reliability, security, and intelligence level of the UAV multi-link communication system, and effectively ensuring the flight safety and mission execution capabilities of the UAV.

[0048] In one embodiment of this invention, the physical layer parameters include real-time signal-to-noise ratio, transmission delay, and available bandwidth. For each heterogeneous communication link, the physical layer parameters are mapped to a virtual vibration displacement sequence, including the following steps: S210, determining virtual mass parameters for each heterogeneous communication link; S220, mapping the available bandwidth of the heterogeneous communication link to a stiffness coefficient and the transmission delay to a damping coefficient, wherein the stiffness coefficient is positively correlated with the available bandwidth and the damping coefficient is positively correlated with the transmission delay; S230, using the reciprocal of the real-time signal-to-noise ratio of the heterogeneous communication link as the excitation force function; S240, constructing a second-order differential equation, which includes the virtual mass parameter, stiffness coefficient, damping coefficient, and excitation force function; S250, iteratively solving the second-order differential equation to generate a virtual vibration displacement sequence for the heterogeneous communication link.

[0049] In this embodiment, determining the virtual quality parameters for each heterogeneous communication link includes the following steps: obtaining the link type identifier of the heterogeneous communication link, which is used to distinguish different communication protocol types; querying the corresponding baseline quality value from a preset link type quality mapping table based on the link type identifier; obtaining the theoretical maximum bandwidth of the heterogeneous communication link and calculating the ratio of the available bandwidth to the theoretical maximum bandwidth of the heterogeneous communication link as the bandwidth utilization rate; calculating the quality correction coefficient based on the bandwidth utilization rate, wherein the higher the bandwidth utilization rate, the larger the quality correction coefficient; and multiplying the baseline quality value by the quality correction coefficient to obtain the virtual quality parameters.

[0050] Virtual mass parameters are determined for each heterogeneous communication link. These parameters represent the inertial characteristics of the link in the virtual vibration system and directly affect the system's response speed and stability to interference. First, the link type identifier of the heterogeneous communication link is obtained. This identifier distinguishes different communication protocol types; for example, 5G cellular networks are identified as TYPE_5G, 2.4GHz data radios as TYPE_RADIO, and satellite communication as TYPE_SAT. The link type identifier is directly read through the communication module's hardware interface or driver layer API, ensuring accurate type identification. After obtaining the identifier, the corresponding baseline mass value is retrieved from a pre-defined link type mass mapping table. This mapping table is pre-established based on the physical characteristics of different communication protocols. For example, the baseline mass value for a 5G link is set to 0.8 kg, reflecting its high-frequency, wide-bandwidth characteristics; the baseline mass value for a 2.4G radio is set to 1.2 kg, reflecting its mid-frequency, medium-bandwidth characteristics; and the baseline mass value for a satellite link is set to 1.5 kg, reflecting its high latency, narrow bandwidth characteristics. The higher the baseline quality value, the greater the response inertia of this type of link, and the slower its response to interference.

[0051] Next, the theoretical maximum bandwidth of the heterogeneous communication links is obtained. This parameter is defined by the communication protocol standard; for example, the theoretical maximum bandwidth of a 5G link is 100Mbps, 2.4G radio is 20Mbps, and satellite links are 5Mbps. The ratio of the currently available bandwidth to the theoretical maximum bandwidth is calculated as the bandwidth utilization rate. ,in Currently available bandwidth, This represents the theoretical maximum bandwidth. For example, if a 5G link currently has a usable bandwidth of 50Mbps and a theoretical maximum bandwidth of 100Mbps, then the bandwidth utilization rate is 0.5. Bandwidth utilization reflects the link's load status and resource usage. A higher utilization rate indicates that the link carries more services, and the system's effective quality should increase accordingly.

[0052] Calculate the quality correction factor based on bandwidth utilization. This formula ensures that the correction factor is 1 when the bandwidth utilization is 0 and 1.5 when the bandwidth utilization is 1, keeping the correction range within a reasonable range. Finally, the baseline quality value is... Multiply by the mass correction factor to obtain the virtual mass parameter. For example, if the baseline mass of a 5G link is 0.8 kg, the bandwidth utilization rate is 0.5, and the correction factor is 1.25, then the virtual mass parameter is 1.0 kg. This step, by introducing differences in link type and dynamic load status, enables the virtual mass parameter to accurately reflect the actual characteristics of different links, laying the foundation for subsequently building an accurate vibration model. This is a key prerequisite for achieving multi-link collaborative monitoring and directly contributes to improving the accuracy of identifying interference from the same source.

[0053] The available bandwidth of heterogeneous communication links is mapped to stiffness coefficients, and transmission delay is mapped to damping coefficients, establishing a quantitative correspondence between communication parameters and vibration system parameters. Stiffness coefficient Characterizes the resilience of a system, corresponding to the ability of a communication link to resist interference and return to normal operation. The mapping relationship is as follows: ,in This is the stiffness mapping coefficient, calibrated based on experimental data, and typically takes a value of 0.1 N / m². The physical meaning of this linear mapping relationship is that the larger the available bandwidth, the stronger the data transmission capability of the link, and the faster it can recover from interference by adding redundant coding, retransmission mechanisms, etc., thus resulting in a larger stiffness coefficient. For example, when the available bandwidth of a 5G link is 50Mbps, the stiffness coefficient is 5N / m; when the bandwidth drops to 30Mbps, the stiffness coefficient drops to 3N / m, and the system's recovery capability weakens accordingly. The positive correlation between the stiffness coefficient and the available bandwidth ensures the physical rationality of the mapping, enabling the virtual vibration system to truly reflect the dynamic characteristics of the communication link.

[0054] Damping coefficient Characterizing the energy dissipation characteristics of a system, corresponding to link response delay and signal attenuation in a communication system. The mapping relationship is as follows: ,in The damping mapping coefficient is 0.05 N·s / (m·ms). The transmission delay is expressed in milliseconds. The physical meaning of this mapping is that the greater the transmission delay, the slower the link response, and the more delayed the system's reaction to interference, exhibiting a stronger damping effect. For example, when the satellite link transmission delay is 200ms, the damping coefficient is 10 N·s / m, much larger than the 1.5 N·s / m of a 5G link with a delay of 30ms, reflecting the high latency characteristics of satellite links. The positive correlation between the damping coefficient and the transmission delay accurately characterizes the impact of delay on the system's dynamic response, causing different types of links to exhibit differentiated dynamic behaviors in the virtual vibration system.

[0055] Through the above mapping, abstract communication parameters are transformed into vibration system parameters with clear physical meaning, providing a parameter basis for constructing a unified mathematical model. This step establishes a bridge between communication quality and vibration characteristics, enabling the use of mature vibration analysis theory to address communication problems. This is a necessary condition for realizing frequency domain coherence analysis and directly supports the quantitative identification capability of co-source interference.

[0056] By using the reciprocal of the real-time signal-to-noise ratio (SNR) of the heterogeneous communication link as the excitation force function, a mapping relationship between interference intensity and system excitation is established. (Real-time SNR) The signal-to-noise ratio (SNR), measured in decibels (dB) in real time by the RF front-end of the communication module, reflects the power ratio of the useful signal to the noise. A higher SNR indicates better communication quality and lower susceptibility to interference; a lower SNR indicates worse communication quality and higher susceptibility to interference. To characterize this interference effect in a vibration system, the reciprocal of the SNR is used as the excitation force, i.e. ,in This is the force scaling factor, used to convert the dimensionless reciprocal of the signal-to-noise ratio (SNR) into an excitation force with a force unit (Newton), typically set to 100 N. The physical meaning of this mapping is that as the SNR decreases, the reciprocal increases, the excitation force increases, indicating that the interference experienced by the link is intensified, driving the virtual vibration system to produce a larger displacement response. For example, when the SNR of a 5G link drops from 25 dB to 15 dB, the excitation force increases from 4 N to 6.67 N, an increase of 66.7%, and the vibration amplitude of the system increases accordingly. When the SNR is extremely low (e.g., 5 dB), the excitation force reaches 20 N, the system enters a forced vibration state, and the displacement increases significantly.

[0057] To avoid mathematical singularities caused by zero or negative signal-to-noise ratios (SNR), a lower limit of 1 dB for SNR is set, corresponding to a maximum excitation force of 100 N. The time-varying characteristics of the excitation force function can reflect the dynamic process of link interference in real time. When an external interference source appears, the SNR of multiple links may decrease synchronously, and the corresponding excitation force increases synchronously, driving the virtual vibration system of each link to generate related response modes. This correlation is the physical basis for subsequent coherence function analysis. This step transforms communication quality indicators into the driving force of the vibration system, establishing a causal relationship between interference and system response. This allows co-source interference to exhibit a coordinated excitation mode in the virtual vibration domain, providing a signal source for identifying multi-link correlations. This is a key technical step for achieving early warning.

[0058] A second-order differential equation is constructed, integrating virtual mass parameters, stiffness coefficients, damping coefficients, and excitation force functions to form a complete mathematical model of the virtual vibration system. The standard form of the second-order differential equation is: in For virtual quality parameters, The damping coefficient is... This is the stiffness coefficient. Let be the excitation force function. For virtual vibration displacement, For vibration velocity, Let be the vibration acceleration. This equation describes a forced damped vibration system. The three terms on the left represent the inertial force, damping force, and elastic restoring force, respectively, while the right side represents the external excitation.

[0059] The physical meaning of the second-order differential equation lies in the vibration of a system driven by an external excitation force. Inertial force hinders the change in acceleration, damping force consumes energy leading to amplitude decay, and elastic force attempts to pull the system back to its equilibrium position. In the mapping of a communication system, this equation describes the dynamic evolution of communication quality under interference: interference (excitation force) leads to a decrease in communication quality (increased displacement), the bandwidth capacity (stiffness) of the link attempts to restore communication quality, delay characteristics (damping) cause a lag in the recovery process, and the load state (quality) of the link affects the response speed.

[0060] For example, for a 5G link, at a certain moment kg, N·s / m, N / m, When N, the equation is The solution to this equation This is a virtual vibration displacement sequence, whose dynamic characteristics are entirely determined by communication parameters. When multiple links are subjected to interference from the same source, the excitation force function of each link... Having similar time-varying patterns leads to displacement sequences for each link. It exhibits characteristics of coordinated change, which manifests as high coherence values ​​in the frequency domain. This step establishes a unified mathematical framework, incorporating heterogeneous communication links of different types and characteristics into the same vibration system model. This allows for the processing of multi-link data using a unified analysis method, thereby achieving quantitative assessment of cross-link correlation and ensuring the computational feasibility of identifying common-source interference.

[0061] The second-order differential equation is solved iteratively to generate a virtual vibration displacement sequence for the heterogeneous communication link. This is because the equation contains a time-varying excitation force function. Furthermore, the system parameters may change dynamically with the communication status, making analytical solutions impossible and requiring numerical methods. The fourth-order Runge-Kutta method (RK4) was chosen for iterative calculation. This method has fourth-order accuracy and is widely used in the numerical solution of ordinary differential equations.

[0062] Specifically, first, the second-order differential equation is transformed into a system of first-order differential equations: Let , Then the system of equations is: Set time step The time interval is set to seconds, consistent with the sampling period of the physical layer parameters, ensuring the accuracy of time discretization. Initial conditions are set to... , This indicates that the system is initially in a static state. At each time step, system parameters are calculated based on the communication parameters at the current time. , , and excitation force Then, the RK4 formula is applied to calculate the state at the next time step. The RK4 iterative formula is as follows: First, calculate the four slopes. , , , Then calculate the weighted average. ,in Let be the right-hand function of the system of differential equations. For example, in At that moment, the initial state of the 5G link was , The system parameters are , , , After RK4 iterative calculation, the following results were obtained. Displacement per second m, velocity m / s. Iterate continuously until the preset time window ends (e.g., 10 seconds, 100 sampling points in total) to generate a complete virtual vibration displacement sequence. This sequence realistically reflects the dynamic process of interference affecting the link throughout the entire time window. The magnitude of the displacement characterizes the interference intensity, and the pattern of displacement variation characterizes the time-frequency characteristics of the interference. This step transforms the theoretical model into a computable numerical sequence, making the abstract communication quality evolution process concrete into analyzable time-series data. This provides a data foundation for subsequent frequency domain transformation and coherence analysis, and determines the sensitivity and accuracy of identifying co-source interference.

[0063] This embodiment realizes a complete mapping process from communication physical layer parameters to virtual vibration displacement sequences, effectively solving the problems of inconsistent data formats and incomparable physical meanings in heterogeneous communication links, and providing a unified analysis framework for multi-link collaborative monitoring. By introducing differences in link type and dynamic load status to determine virtual quality parameters, different types of links have differentiated dynamic characteristics in the same model, accurately reflecting the essential differences between heterogeneous links such as 5G, 2.4G radio, and satellite, which is a prerequisite for accurate modeling. Establishing a quantitative mapping relationship transforms communication capability into system resilience and latency characteristics into response sluggishness, so that changes in communication parameters can directly drive changes in the dynamic behavior of the vibration system. Using the reciprocal of the signal-to-noise ratio as the excitation force, a causal relationship between interference and system excitation is established, allowing co-source interference to exhibit a collaborative excitation mode in multiple links, serving as the signal source for coherence analysis. The constructed second-order differential equation integrates all mapping parameters, forming a complete mathematical model that retains the individual characteristics of each link while providing a unified analysis interface. Transforming the theoretical model into a computable time series lays the data foundation for subsequent frequency domain analysis, effectively enabling early identification of interference from the same source, solving the technical problem that traditional single-link monitoring methods cannot detect multi-link collaborative decay, and significantly improving the security and reliability of UAV communication systems.

[0064] In one embodiment of this invention, the heterogeneous communication link includes a first heterogeneous communication link and a second heterogeneous communication link. The cross-power spectral density between the spectral functions of the different heterogeneous communication links is calculated, and the coherence function and its maximum coherence value are determined based on the cross-power spectral density. This includes the following steps: S310, performing windowed Fourier transforms on the virtual vibration displacement sequences of the first and second heterogeneous communication links respectively to obtain a first spectral function and a second spectral function; S320, calculating the cross-power spectral density based on the first and second spectral functions; S330, calculating the first auto-power spectral density of the first spectral function and the second auto-power spectral density of the second spectral function respectively; S340, calculating the coherence function based on the cross-power spectral density, the first auto-power spectral density, and the second auto-power spectral density; S350, traversing the coherence values ​​of the coherence function at different frequency points to extract the maximum coherence value.

[0065] Windowed Fourier transforms are performed on the virtual vibration displacement sequences of the first and second heterogeneous communication links to obtain their corresponding spectral functions. The first heterogeneous communication link can be a 5G cellular network link, and the second heterogeneous communication link can be a 2.4GHz data radio link. The two links have generated their respective virtual vibration displacement sequences through the aforementioned steps, denoted as... and The sequence length is 100 sampling points, corresponding to a 10-second time window.

[0066] To reduce spectral leakage, the sequence needs to be windowed before performing the Fourier transform. The Hanning window function is chosen, and its expression is: ,in The sampling point number (0 to 99), The window length is 100. The Hanning window smoothly transitions to zero at both ends of the time domain, effectively suppressing spectral leakage while maintaining good frequency resolution. The windowed sequence is... and .

[0067] Performing a Fast Fourier Transform (FFT) on the windowed sequence yields the spectral function. and The FFT transforms a time-domain signal into the frequency domain. The spectral function is in complex form and contains amplitude and phase information, with a frequency resolution of [missing value]. Hz, of which The time window length is 10 seconds. The frequency range of the spectrum function is 0 to 5 Hz (half of the sampling frequency of 10 Hz, i.e., the Nyquist frequency), with a total of 51 frequency points. For example, the spectrum function of a 5G link. The value at 0.5Hz , indicating that the amplitude of this frequency component is Phase is This step transforms the virtual vibration displacement sequence in the time domain to the frequency domain, allowing analysis of the interference patterns of the links from the perspective of frequency characteristics. When multiple links are subjected to interference from the same source, the interference source often has specific frequency characteristics (such as the scanning frequency of sweep interference and the repetition frequency of pulse interference). These characteristics may be masked by noise in the time domain, but they will appear as obvious peaks in the frequency domain. Frequency domain analysis can more clearly identify the frequency fingerprint of the interference from the same source.

[0068] The cross-power spectral density is calculated based on the first and second spectral functions to quantify the correlation between the two links in the frequency domain. The cross-power spectral density is defined as the frequency domain representation of the cross-correlation of the two spectral functions, and its calculation formula is as follows: ,in express The conjugate of the complex number. For complex numbers Its conjugate complex number is Cross-power spectral density It is also a plural number, which actually reflects the frequency of the two links. The in-phase correlation at the point, the imaginary part reflects the orthogonal correlation, and the modulus It reflects the overall correlation strength.

[0069] For example, at a frequency of 0.5 Hz, , ,but Cross-power spectral density The modulus is The cross-power spectral density was calculated for all 51 frequency points to obtain the complete cross-power spectral density function. .

[0070] The physical meaning of cross-power spectral density (CPSD) is that when the signal components of two links at a certain frequency are highly correlated (similar amplitudes and phases), the magnitude of the CPSD at that frequency is large; when the signal components of two links at a certain frequency are uncorrelated or out of phase, the magnitude of the CPSD is small. This step establishes a quantitative index of cross-link frequency domain correlation, providing a mathematical tool for identifying co-source interference. When external interference sources act on multiple links simultaneously, the spectral functions of each link will exhibit similar peak characteristics near the interference frequency, leading to a significant increase in the CPSD within that frequency range. By analyzing the distribution pattern of the CPSD, the frequency characteristics of co-source interference can be identified.

[0071] The first self-power spectral density of the first spectral function and the second self-power spectral density of the second spectral function are calculated respectively to provide a benchmark for the normalized coherence function. The self-power spectral density represents the energy distribution of a single link at various frequency points and is the product of the spectral function and its own conjugate. The formula for calculating the self-power spectral density of the first link is as follows: The self-power spectral density of the second link is Since the square of the magnitude of the spectral function is equal to its product with its conjugate, the self-power spectral density is real and non-negative.

[0072] For example, at a frequency of 0.5 Hz, ,but Similarly, , The complete self-power spectral density function was obtained by calculating for all 51 frequency points. and The peak value of the self-power spectral density corresponds to the main frequency component of the link being interfered with; the higher the peak value, the more concentrated the energy at that frequency. For example, if a 5G link exhibits a peak value of self-power spectral density at 1.2Hz, it indicates that the link is significantly interfered with near that frequency or has an inherent oscillation mode. This step provides a normalized denominator for the coherence function calculation, enabling the coherence function to eliminate the influence of the energy levels of each link and achieve standardized comparison of correlations between different links. Without normalization, links with higher energy levels would dominate the cross-power spectral density value, masking the true degree of correlation. By dividing by their respective self-power spectral densities, the coherence function value is limited to between 0 and 1, possessing clear statistical significance and facilitating the setting of a unified judgment threshold.

[0073] The coherence function is calculated based on the cross-power spectral density, the first self-power spectral density, and the second self-power spectral density, yielding a normalized correlation index. The calculation of the coherence function involves three sub-steps: first, calculating the square of the modulus of the cross-power spectral density... This value indicates the frequency of the two links. The energy of the correlation is calculated; then the product of the power spectral densities of the two links is calculated. This value represents the geometric mean of the energy of the two links at that frequency; finally, dividing the former by the latter yields the coherence function. .

[0074] For example, at a frequency of 0.5 Hz, it is known , , ,but , coherence function The result indicates that the two links are completely coherent at 0.5 Hz and may be affected by the same interference source.

[0075] The range of values ​​for the coherence function is strictly limited to between 0 and 1. When the signal is fully coherent, the signal components of the two links at that frequency are linearly correlated; when... The time interval indicates complete incoherence, meaning the signal components of the two links at that frequency are independent of each other. The complete coherence function is obtained by calculating for all 51 frequency points. This step transforms cross-link correlation into a standardized statistical indicator, eliminating the influence of differences in energy levels across links, and allowing a uniform threshold to determine the existence of co-source interference. The peak position of the coherence function reveals the frequency characteristics of co-source interference, and the peak height reflects the degree of synchronization of the interference, enabling quantitative identification of co-source interference.

[0076] The coherence function values ​​at different frequency points are iterated through, and the maximum coherence value is extracted as a comprehensive criterion for multi-link cooperative fading. Coherence function values ​​are read one by one from 51 frequency points from 0Hz to 5Hz. Find the maximum value through comparison operations. For example, if the coherence function reaches its maximum value of 0.85 at 1.2 Hz, then The corresponding frequency is 1.2Hz.

[0077] The maximum coherence value and its corresponding frequency contain important diagnostic information: the magnitude of the maximum coherence value reflects the strongest correlation between two links; if this value is significantly higher than the statistical expectation of independent links (usually below 0.3), it indicates the presence of co-source interference. The frequency corresponding to the maximum coherence value reveals the frequency characteristics of the interference source, which can be used for subsequent interference source localization and avoidance strategy formulation. In practical applications, to improve robustness, the first three peaks of the coherence function and their frequencies can be extracted to comprehensively determine the interference mode. For example, if the coherence function shows peaks at 1.2Hz, 2.4Hz, and 3.6Hz, and the peaks are evenly distributed, it can be inferred that the interference source is periodic pulse interference with a fundamental frequency of 1.2Hz. This step condenses the frequency domain coherence analysis results into a single decision indicator, simplifying the subsequent judgment logic and enabling the system to evaluate the cooperative status of multiple links in real time and efficiently. As a sensitive indicator of co-source interference, the maximum coherence value can issue an early warning before the single-link indicator triggers a traditional alarm, buying valuable time for taking proactive avoidance measures.

[0078] This embodiment implements a complete calculation process from virtual vibration displacement sequence to maximum coherence value, enabling quantitative assessment of multi-link collaborative monitoring and effectively solving the technical problem of difficulty in quantifying the correlation between heterogeneous communication links. By using windowed Fourier transform to convert the time-domain signal to the frequency domain, the limitations of noise masking effects in time-domain analysis are overcome. The calculated cross-power spectral density establishes a quantitative index of cross-link frequency domain correlation, allowing the collaborative characteristics of co-source interference to be mathematically expressed. The calculated self-power spectral density provides a benchmark for normalization, ensuring the statistical significance and comparability of the coherence function. The calculated coherence function transforms the correlation into a standardized index of 0 to 1, eliminating the influence of link energy differences and allowing for the use of a unified threshold to determine co-source interference. The extracted maximum coherence value condenses the frequency domain analysis results into a single decision index, simplifying the real-time judgment logic and improving the system response speed. Compared to traditional single-link independent monitoring methods, this scheme can identify the collaborative decay trend of multiple links in the early stages of slight fluctuations in communication indicators, realizing a shift from passive response to proactive early warning, and significantly improving the safety and reliability of UAV communication systems.

[0079] In one embodiment of this example, the coherence function is calculated based on the cross-power spectral density, the first self-power spectral density, and the second self-power spectral density, including the following steps: S410, calculating the square of the modulus of the cross-power spectral density; S420, calculating the product of the first self-power spectral density and the second self-power spectral density; S430, dividing the square of the modulus of the cross-power spectral density by the product to obtain the coherence function.

[0080] Calculate the square of the modulus of the cross-power spectral density to obtain an energy characterization of the correlation between the two links at each frequency point. Cross-power spectral density It is in complex form and can be represented as ,in For the real part, The imaginary part is denoted by . The modulus of a complex number is defined as . The square of the modulus is This calculation process eliminates phase information, retaining only the intensity of the correlation.

[0081] For example, at a frequency of 1.2 Hz, if the cross-power spectral density is... Then the real part imaginary part The square of the modulus is Perform this calculation for all 51 frequency points to obtain the cross-power spectral density modulus square sequence. .

[0082] The modulus square operation transforms the correlation in the complex domain into an energy metric in the real domain; a larger value indicates a stronger correlation between the signal components of the two links at that frequency. This step provides a standardized energy index for the numerator of the coherence function, making the correlations at different frequencies and times comparable. When co-source interference affects multiple links, the squared modulus of the cross-power spectral density at the interference frequency increases significantly. This characteristic is a key signal for identifying co-source interference. Through this calculation, the abstract frequency domain correlation is transformed into a directly comparable value, laying the foundation for subsequent normalization processing.

[0083] Calculate the product of the first and second self-power spectral densities to obtain the denominator of the normalized coherence function. Self-power spectral density of the first link. Self-power spectral density of the second link All are real numbers and non-negative, their product This indicates that the two links are in frequency The geometric mean of their respective energies. For example, at a frequency of 1.2 Hz, if... , Then the product The power spectral density product sequence was obtained by calculating for all 51 frequency points. .

[0084] This product provides an energy benchmark for the correlation: if two links have high energy at a certain frequency, even if the cross-power spectral density is high, it may only be due to the strength of their individual signals, rather than a true correlation. By dividing by the product of their own power spectral densities, the influence of the energy levels of each link can be eliminated, and the pure degree of correlation can be extracted. This step establishes a normalized benchmark, enabling the coherence function to reflect the true degree of linear correlation without being disturbed by differences in link energy.

[0085] In UAV multi-link systems, the transmit power, antenna gain, and transmission loss of different types of links vary greatly, resulting in signal energy levels that differ by several orders of magnitude. Without normalization, high-energy links will dominate the correlation calculation results, masking the cooperative decay characteristics of weak links. This normalization step ensures that all links have equal weight in the coherence assessment, achieving fair comparison across heterogeneous links.

[0086] The coherence function is obtained by dividing the square of the magnitude of the cross-power spectral density by the product of the self-power spectral densities. The formula for calculating the coherence function is as follows: This operation is performed independently at each frequency point; for example, at the 1.2Hz frequency point, it is known that... , Then the coherence function Theoretically, the coherence function should be between 0 and 1. If the calculated result is slightly greater than 1 (e.g., 1.062), it is usually due to numerical calculation errors or statistical fluctuations caused by finite sampling. In practical applications, it can be truncated to 1. The complete coherence function is obtained by calculating it separately for all 51 frequency points. .

[0087] The coherence function has a clear statistical meaning: when When, it indicates that the two links are at the same frequency. At a point where they are perfectly linearly correlated, the signal of one link can be completely predicted by the signal of the other link; when... When, it means that the two links are completely independent at that frequency and are uncorrelated; when When the value is close to 1, it indicates a partial correlation; the closer the value is to 1, the stronger the correlation.

[0088] This step condenses complex frequency domain correlation analysis into an intuitive normalized index, enabling the determination of co-source interference based on a unified threshold. In traditional single-link monitoring methods, the signal-to-noise ratio, bit error rate, and other indicators of each link have different dimensions and numerical ranges, making it difficult to establish a unified judgment standard. However, the coherence function, as a dimensionless normalized index, has a fixed value range and clear physical meaning. A unified threshold for determining co-source interference (e.g., 0.7) can be set. When the coherence function exceeds this threshold, co-source interference can be determined to exist. This method achieves quantitative and automated co-source interference identification, overcoming the shortcomings of traditional methods that rely on human experience, are highly subjective, and have low accuracy.

[0089] This embodiment achieves accurate solution of the coherence function, establishes a quantitative evaluation index for multi-link cooperative status, and effectively solves the problem of normalizing the correlation of heterogeneous communication links. By calculating the square of the modulus of the cross-power spectral density, the correlation in the complex domain is transformed into an energy metric in the real domain, eliminating the interference of phase information and extracting the intensity characteristics of the correlation, making the correlation at different frequency points comparable. By calculating the product of the self-power spectral densities, a normalization benchmark is established, eliminating the influence of differences in the energy levels of each link and ensuring fairness in the correlation evaluation of links of different types and power levels. The coherence function is obtained through division, transforming the correlation into a standardized index of 0 to 1, allowing the use of a unified threshold to determine co-source interference, simplifying the decision-making logic, and improving the real-time performance and reliability of the system. Compared with the traditional single-link independent monitoring method, this scheme achieves fair comparison across links through normalization processing, reveals the frequency fingerprint of co-source interference through frequency domain analysis, and quantifies the degree of cooperative decay through the coherence function. It can accurately identify co-source interference in the early stages of slight fluctuations in communication indicators, buying valuable time for taking proactive avoidance measures and significantly improving the security and reliability of UAV multi-link communication systems.

[0090] In one embodiment of this invention, the flight attitude data includes three-axis acceleration and three-axis angular velocity. Dynamically adjusting a preset threshold for determining common-source interference based on the flight attitude data includes the following steps: S510, calculating the total acceleration of the UAV based on the three-axis acceleration; S520, calculating the maneuver intensity factor based on the total acceleration; S530, calculating the angular velocity magnitude of the UAV based on the three-axis angular velocity; S540, calculating the attitude change rate based on the angular velocity magnitude; S550, calculating the comprehensive maneuver index based on the maneuver intensity factor and the attitude change rate; S560, calculating the threshold adjustment coefficient based on the comprehensive maneuver index, wherein a higher comprehensive maneuver index results in a larger threshold adjustment coefficient; S570, multiplying the preset threshold for determining common-source interference by the threshold adjustment coefficient to obtain the adjusted threshold for determining common-source interference.

[0091] The total acceleration of the UAV is calculated based on the three-axis acceleration, quantifying the intensity of the UAV's maneuvers. The three-axis acceleration is obtained in real time by the onboard inertial measurement unit (IMU), including the acceleration along the X-axis (nose direction) of the body coordinate system. Y-axis (right side of the fuselage) acceleration and acceleration along the Z-axis (below the fuselage) The units are all m / s². The sampling frequency of the IMU is usually 100Hz, which is different from the 10Hz sampling frequency of the communication parameters. Time alignment processing is required. Nearest neighbor interpolation or linear interpolation methods can be used to downsample the IMU data to 10Hz.

[0092] The formula for calculating the total acceleration of an organism is: This formula, based on the principle of calculating the magnitude of a vector in three-dimensional space, yields the Euclidean norm of the acceleration vector. For example, at a certain moment, the acceleration vector is measured... m / s² m / s² m / s², then the total acceleration of the body m / s². When the UAV is in stable flight, the three-axis acceleration is mainly composed of the gravitational acceleration component, with a total acceleration close to 9.8 m / s². When the UAV performs maneuvers such as sharp turns, climbs, or dives, additional inertial acceleration is generated, significantly increasing the total acceleration to 15 m / s² or even higher. This step establishes the correlation between flight maneuvers and communication quality fluctuations. In actual flight, UAV maneuvers cause physical effects such as changes in antenna pointing, signal propagation paths, and Doppler shifts. These effects simultaneously affect the signal quality of multiple communication links, leading to coordinated fluctuations in the communication parameters of each link. This fluctuation is superficially similar to the coordinated fading caused by co-source interference, but the physical mechanisms are completely different. By calculating the total acceleration of the UAV, the severity of flight maneuvers can be quantified, providing a basis for distinguishing between normal maneuver fluctuations and abnormal interference fading.

[0093] Based on the total acceleration of the aircraft, a normalized relationship between acceleration and maneuverability is established using the calculated dynamic intensity factor. The maneuverability intensity factor is defined as the ratio of the total acceleration of the aircraft to the acceleration due to gravity, and the calculation formula is as follows: ,in m / s² represents standard gravitational acceleration. This normalization process makes the maneuverability factor a dimensionless parameter, facilitating comprehensive calculations with other indicators. For example, when the total acceleration of the aircraft is 7.07 m / s², the maneuverability factor... When the total acceleration of the aircraft is 14.7 m / s², the maneuverability factor is... .

[0094] The maneuverability intensity factor characterizes the overload factor that a UAV experiences. When, it means the drone is only affected by gravity and is in a state of uniform linear flight or hovering; when At this time, it indicates that the drone is performing maneuvers such as acceleration and turning; the higher the overload multiple, the more violent the maneuver. In theory, this indicates that the drone is in a state of weightlessness or deceleration, but this rarely occurs in practical applications. This step converts absolute acceleration into relative maneuver intensity, making the degree of maneuverability comparable under different flight conditions.

[0095] In UAV communication systems, maneuvers can cause changes in the effective gain of the antenna pattern, rapid changes in the signal propagation path, and enhanced Doppler effect. These phenomena simultaneously affect multiple links, causing synchronous fluctuations in parameters such as signal-to-noise ratio and transmission delay across each link. This leads to a correlation in the virtual vibration displacement sequence, resulting in an increase in the coherence function value. If flight attitude factors are not considered, this correlation can be misjudged as interference from the same source, causing the system to perform unnecessary evasive maneuvers and impacting mission efficiency. By introducing a maneuver intensity factor, the impact of flight maneuvers on communication quality can be quantitatively assessed, providing a quantitative basis for dynamically adjusting the judgment threshold.

[0096] The angular velocity magnitude of the UAV is calculated based on the three-axis angular velocities, quantifying the rate of attitude change of the UAV. The three-axis angular velocities are also obtained by IMU measurement, including the roll angular velocity around the X-axis. Pitch angular velocity around the Y-axis and yaw rate about the Z-axis All units are rad / s. The formula for calculating the magnitude of angular velocity is: For example, measured at a certain moment rad / s rad / s rad / s, then the magnitude of angular velocity rad / s.

[0097] When a drone is in stable flight, its three-axis angular velocity is close to zero, and its angular velocity magnitude is very small (typically less than 0.05 rad / s). When the drone performs rapid turns, rolls, or other attitude adjustment maneuvers, its angular velocity increases significantly, reaching over 1 rad / s. The angular velocity magnitude directly reflects the rate of change of the drone's attitude and is an important indicator for evaluating antenna pointing stability. This step introduces the dimension of attitude change rate, supplementing the shortcomings of the acceleration indicator.

[0098] In certain flight scenarios, a drone may maintain a low linear acceleration but perform rapid attitude adjustments (such as rotation in place or yaw while hovering). In these situations, the total acceleration of the drone remains relatively constant, but the angular velocity is very high. Rapid attitude changes cause a rapid deflection of the antenna's pointing direction, resulting in synchronous changes in signal strength, multipath effects, and line-of-sight conditions across various links, leading to coordinated fluctuations in communication quality. Considering only acceleration will overlook the impact of these attitude maneuvers on communication, leading to misjudgments. By calculating the magnitude of the angular velocity, the drone's maneuvering characteristics can be comprehensively captured, providing more complete information for accurately distinguishing between normal maneuvers and abnormal interference.

[0099] The attitude change rate is calculated based on the angular velocity magnitude, establishing a normalized relationship between angular velocity and the degree of attitude change. The attitude change rate is defined as the ratio of the angular velocity magnitude to the reference angular velocity, and the calculation formula is as follows: ,in The reference angular velocity is set to 0.1 rad / s. This reference angular velocity is selected based on the typical angular velocity level during normal UAV flight, representing the boundary between stable and maneuvering flight. For example, when the angular velocity magnitude is 0.25 rad / s, the attitude change rate... When the angular velocity modulus is 0.05 rad / s, the rate of change of attitude .

[0100] The rate of attitude change characterizes the drasticness of attitude change. When the drone's attitude changes slowly, it is in a stable flight state; when This indicates that the UAV is performing a rapid attitude adjustment; the greater the rate of attitude change, the more drastic the adjustment. This step converts the absolute angular velocity into the relative degree of attitude change, making attitude maneuvers under different flight conditions comparable.

[0101] In multi-link communication systems, the installation location and radiation pattern characteristics of an antenna determine its sensitivity to attitude changes. For example, omnidirectional antennas are insensitive to attitude changes, while directional antennas are highly sensitive. When a UAV performs rapid attitude adjustments, the antenna gain change patterns of different links may be similar (e.g., all undergoing a process of deviating from the direction of maximum gain), leading to a synchronous decrease in the received signal strength of each link. This, in turn, causes a coordinated change in the virtual vibration displacement sequence, resulting in an increase in the coherence function value. By introducing the attitude change rate, the impact of attitude maneuvers on communication quality can be quantitatively assessed, providing an attitude dimension basis for dynamically adjusting the judgment threshold. This complements the maneuver intensity factor, jointly constructing a comprehensive flight state assessment system.

[0102] A comprehensive maneuverability index is calculated based on the maneuver intensity factor and attitude change rate to establish a comprehensive evaluation index for the impact of flight maneuvers on communication. The comprehensive maneuverability index is calculated using a weighted summation method, and the formula is as follows: ,in The weights of the maneuverability factor, As the weight of the attitude change rate, and Based on experimental data and engineering experience, linear acceleration typically has a greater impact on communication quality than angular velocity; therefore, it is set... , For example, when the maneuver intensity factor is 1.5 and the attitude change rate is 2.5, the comprehensive maneuver index is... .

[0103] Theoretically, the comprehensive mobility index ranges from 0 to infinity, but in practical applications it is usually between 0 and 3. When indicates stable flight, When indicates moderate maneuverability, The time indicates a severe maneuver. This step integrates multi-dimensional flight state information into a single comprehensive index, simplifying the subsequent threshold adjustment logic. In actual flight, linear acceleration and angular velocity often change simultaneously; for example, during a sharp turn, there is both centrifugal acceleration and yaw rate. Processing these two dimensions separately would increase algorithm complexity and computational burden. Through weighted fusion, a single index can comprehensively reflect the severity of flight maneuvers. The weighting coefficients are set based on the relative importance of the two types of maneuvers to communication, ensuring the physical rationality of the comprehensive index.

[0104] A higher overall maneuverability index indicates more intense flight maneuvers, greater normal fluctuations in communication quality, and a higher probability of coordinated fluctuations across links. In this case, the threshold for identifying common-source interference should be increased to avoid misjudging normal maneuvering fluctuations as common-source interference. Conversely, when the overall maneuverability index is low, communication quality should remain stable. If coordinated fluctuations occur, it is likely to be common-source interference. In this case, a lower threshold should be maintained to improve detection sensitivity.

[0105] A threshold adjustment coefficient is calculated based on the comprehensive mobility index to establish a quantitative relationship between the degree of mobility and the threshold adjustment range. The threshold adjustment coefficient is calculated using a linear mapping method, and the formula is as follows: ,in To adjust the sensitivity coefficient, a value of 0.5 was chosen. This coefficient was selected based on statistical analysis of a large amount of flight test data, ensuring that the adjustment range effectively avoids misjudgment without excessively reducing detection sensitivity. For example, when the comprehensive maneuverability index is 0 (stable flight), the threshold adjustment coefficient... This indicates that the threshold is not adjusted; when the comprehensive maneuver index is 1.9 (aggressive maneuver), the threshold adjustment coefficient is [value missing]. This indicates that the threshold will be increased by 95%.

[0106] The threshold adjustment coefficient ranges from 1 to infinity, and in practical applications it is usually between 1 and 2.5. When the threshold remains unchanged, The threshold is increased, with the increase proportional to the degree of maneuverability. This step establishes an automated mapping mechanism from flight status to threshold adjustment, enabling the monitoring system to adapt.

[0107] Traditional fixed-threshold monitoring methods face a dilemma in threshold setting: if the threshold is set too low, normal fluctuations may be misjudged as interference during maneuvering flight, leading to frequent false alarms and unnecessary evasive maneuvers, thus affecting mission efficiency; if the threshold is set too high, insufficient detection sensitivity may result in missing genuine interference from the same source during stable flight, leading to false alarms and threatening flight safety. By introducing a dynamic threshold adjustment mechanism, the detection sensitivity can be automatically adjusted according to the real-time flight status, maintaining high sensitivity during stable flight and appropriately reducing sensitivity during maneuvering flight, achieving a dynamic balance between sensitivity and robustness.

[0108] The adjusted same-source interference threshold is obtained by multiplying the preset same-source interference determination threshold by the threshold adjustment coefficient. (Preset same-source interference determination threshold) Based on statistical theory and experimental data, the threshold is typically set to 0.7. This threshold indicates that, under the independent link assumption, the probability of the coherence function exceeding this value is less than 1%, thus possessing statistical significance. The adjusted threshold calculation formula is as follows: For example, when the preset threshold is 0.7 and the threshold adjustment coefficient is 1.95, the adjusted threshold... Since the theoretical maximum value of the coherence function is 1, when the adjusted threshold exceeds 1, it is truncated to 1 in practical applications. This means that even if the coherence function reaches its maximum value under this flight condition, it is not considered as interference from the same source, and fault detection relies entirely on single-link indicators. When the overall maneuverability index is low, such as... The threshold adjustment coefficient is 1.1, and the adjusted threshold is 0.77, which is only slightly higher than the preset threshold, maintaining a high detection sensitivity.

[0109] This step completes the closed-loop control from flight status perception to monitoring parameter adjustment, enabling the monitoring system to adapt to the flight environment. The adjusted threshold is directly used for subsequent communication status determination. Only when the maximum coherence value exceeds the adjusted threshold is the regional electromagnetic environment considered deteriorated; otherwise, it is considered a single-link failure or normal communication status. This dynamic threshold mechanism effectively solves the limitations of traditional fixed threshold methods, enabling the monitoring system to adapt to the complex and ever-changing flight states of UAVs. While ensuring detection accuracy, it reduces the false positive rate, significantly improving the system's practicality and reliability.

[0110] This embodiment implements a dynamic adjustment mechanism for the threshold of co-source interference judgment based on flight attitude data, effectively solving the technical problem of difficulty in distinguishing between normal fluctuations in communication quality caused by flight maneuvers and abnormal degradation caused by co-source interference. This mechanism establishes an adaptive correlation between flight state and monitoring parameters, enabling the monitoring system to intelligently adjust detection sensitivity according to real-time flight status. The impact of linear maneuvers on communication is quantified by calculating the total acceleration and maneuver intensity factor; the impact of attitude maneuvers on communication is quantified by calculating the angular velocity magnitude and attitude change rate; a comprehensive maneuver index is obtained through weighted fusion, establishing a comprehensive flight state evaluation system; and dynamic adjustment of judgment parameters is achieved by calculating threshold adjustment coefficients and applying them to preset thresholds. Compared to traditional fixed threshold methods, this dynamic adjustment mechanism maintains high detection sensitivity during stable flight, enabling timely detection of early signs of co-source interference; and appropriately reduces sensitivity during maneuvering flight to avoid misjudging normal fluctuations as interference, reducing false alarm rates and significantly improving the accuracy and reliability of the monitoring system. This provides reliable technical support for the safe operation of UAVs in complex flight and electromagnetic environments.

[0111] In one embodiment of this example, a communication status monitoring conclusion is generated based on the maximum coherence value, the adjusted same-source interference judgment threshold, and the amplitude, including the following steps: S610, if the maximum coherence value exceeds the adjusted same-source interference judgment threshold, and the amplitude of the virtual vibration displacement sequence of at least two heterogeneous communication links exceeds the preset amplitude threshold, the communication status monitoring conclusion is determined to be a deterioration of the regional electromagnetic environment; S620, if the maximum coherence value does not exceed the same-source interference judgment threshold, and the amplitude of the virtual vibration displacement sequence of a single heterogeneous communication link exceeds the preset amplitude threshold, the communication status monitoring conclusion is determined to be a single-link failure; S630, if the maximum coherence value does not exceed the same-source interference judgment threshold, and the amplitude of the virtual vibration displacement sequence of all heterogeneous communication links does not exceed the preset amplitude threshold, the communication status monitoring conclusion is determined to be normal communication status.

[0112] If the maximum coherence value exceeds the adjusted threshold for determining co-source interference, and the amplitudes of the virtual vibration displacement sequences of at least two heterogeneous communication links all exceed a preset amplitude threshold, the communication status monitoring conclusion is determined to be a deterioration of the regional electromagnetic environment. This judgment logic employs a dual-condition judgment mechanism to ensure the accuracy of the diagnostic conclusion. First, the maximum coherence value is compared... The adjusted threshold for determining co-source interference ,like This indicates that multiple links exhibit significant correlation in the frequency domain, satisfying the necessary condition for co-source interference. For example, when the maximum coherence value is 0.88 and the adjusted threshold is 0.77, the first condition is met.

[0113] Next, the amplitude of the virtual vibration displacement sequence of each link is checked, and a preset amplitude threshold is used. Typically set at 1.0m, this threshold is determined based on statistical analysis of extensive experimental data and represents the dividing point between normal and abnormal communication quality. The amplitude of all links is examined; if the amplitude of at least two links exceeds 1.0m (e.g., 1.3m for 5G links and 1.5m for 2.4G links), then the second condition is met. When both conditions are met simultaneously, it is determined that the regional electromagnetic environment has deteriorated. This conclusion indicates that multiple heterogeneous links are affected by the same external interference source, resulting in a coordinated decline in communication quality that has reached a significant level. This step establishes a dual verification mechanism for co-source interference, avoiding the one-sidedness of judging by a single indicator. Relying solely on coherence values ​​may misjudge coordinated fluctuations caused by flight maneuvers as interference, and relying solely on amplitude determination cannot distinguish between co-source interference and the superposition effect of multiple independent interferences. By combining coherence analysis and amplitude detection, both the coordinated characteristics of multiple links are ensured, and the severity of fading is verified, achieving accurate identification of co-source interference.

[0114] If the maximum coherence value does not exceed the threshold for determining common-source interference, and the amplitude of the virtual vibration displacement sequence of a single heterogeneous communication link exceeds a preset amplitude threshold, the communication status monitoring conclusion is determined to be a single-link fault. This determination logic is used to identify independent faults within the link itself. First, it determines whether the maximum coherence value does not exceed the threshold, i.e. This condition indicates that there is no significant correlation between the links in the frequency domain, and the communication quality changes of each link are relatively independent. For example, the first condition is met when the maximum coherence value is 0.65 and the adjusted threshold is 0.77.

[0115] Next, the amplitude of each link is checked. If only a single link's amplitude exceeds the preset amplitude threshold (e.g., 1.4m for a 5G link, 0.7m for a 2.4G link, and 0.6m for a satellite link), then the second condition is met. When both conditions are met simultaneously, it is determined to be a single-link failure. This conclusion indicates that the faulty link itself has a problem, which could be a hardware failure (e.g., a damaged power amplifier or broken antenna), local interference (e.g., narrowband interference in the link's frequency band), or a link configuration error, while other links are functioning normally. This step achieves accurate fault location, avoiding misjudgments and unnecessary handling of healthy links. In traditional single-link independent monitoring methods, the fault determination of each link is independent, making cross-verification using multi-link information impossible and susceptible to misjudgments due to transient noise interference. By introducing coherence analysis, the independence of the fault can be verified. When the coherence value is low, it is confirmed that the abnormality of the link is an isolated event rather than a systemic problem. At this time, a link switching strategy should be implemented to migrate the business to a healthy link, rather than implementing global response measures such as location avoidance. This accurate fault classification capability significantly improves the pertinence and effectiveness of the response strategy.

[0116] If the maximum coherence value does not exceed the threshold for determining common-source interference, and the amplitude of the virtual vibration displacement sequence of all heterogeneous communication links does not exceed the preset amplitude threshold, the communication status monitoring conclusion is determined to be that the communication status is normal. This determination logic is used to confirm that the system is in a healthy operating state. First, it determines whether the maximum coherence value does not exceed the threshold, i.e. This indicates that there is no abnormal collaborative correlation among the links. For example, when the maximum coherence value is 0.60 and the adjusted threshold is 0.77, the first condition is met.

[0117] Next, check the amplitude of all links. If the amplitude of all links does not exceed the preset amplitude threshold (e.g., 0.8m for 5G links, 0.7m for 2.4G links, and 0.6m for satellite links), all less than 1.0m, then the second condition is met. When both conditions are met, the communication status is considered normal. This conclusion indicates that the communication quality of all links is within the normal range, there is no significant interference or fault, and the current configuration can continue to be maintained.

[0118] This step establishes clear criteria for determining normal operation, avoiding frequent alarms and unnecessary interventions caused by oversensitivity. In practical applications, communication parameters can fluctuate randomly due to factors such as environmental noise and channel fading. If the monitoring system is overly sensitive to minute fluctuations, it will lead to frequent false alarms, increasing the system load and interfering with normal task execution. By setting reasonable amplitude and coherence thresholds, normal fluctuations and abnormal fading are clearly distinguished. Alarms and countermeasures are only triggered when the indicators exceed the thresholds, achieving a balance between monitoring sensitivity and system stability.

[0119] This embodiment implements a communication status classification and judgment mechanism based on multi-dimensional indicators, establishing a complete mapping relationship from monitoring data to diagnostic conclusions. This effectively solves the problem of accurate classification of communication anomalies, providing accurate decision-making basis for subsequent targeted response strategies. By verifying the deterioration of the regional electromagnetic environment through dual-condition verification, it uses both coherence values ​​to judge the coordination of multiple links and amplitude to judge the severity of attenuation, achieving accurate identification of interference from the same source. By combining coherence values ​​and amplitude, it identifies single-link faults, achieving accurate fault location, avoiding misjudgments of healthy links, and providing accurate triggering conditions for link switching strategies. By confirming normal communication status through an elimination method, it establishes a judgment standard for normal operation, avoiding frequent false alarms caused by oversensitivity and ensuring stable system operation. The entire judgment mechanism adopts a hierarchical decision-making logic: first, it judges the correlation between links based on coherence values; then, it judges the absolute level of communication quality based on amplitude; and finally, it combines the two types of information to draw a diagnostic conclusion. The logic is clear and the judgment is accurate. Compared with traditional single-link independent monitoring methods, this hierarchical judgment mechanism can accurately distinguish between three situations: interference from the same source, single-link failure, and normal state. The diagnostic accuracy is greatly improved, significantly enhancing the intelligence level and practical value of the monitoring system, and providing a solid technical guarantee for the safe and reliable operation of UAV communication systems.

[0120] In one embodiment of this example, based on the communication status monitoring conclusion, a corresponding link control strategy is executed, including the following steps: S710, if the communication status monitoring conclusion is that the regional electromagnetic environment has deteriorated, the link switching strategy is prohibited, and a location avoidance strategy is executed; S720, if the communication status monitoring conclusion is that a single link has failed, the link switching strategy is executed to switch the communication service to a healthy link; S730, if the communication status monitoring conclusion is that the communication status is normal, a maintenance strategy is executed to maintain the current communication configuration.

[0121] If communication status monitoring indicates a deterioration in the regional electromagnetic environment, link switching should be prohibited, and a location avoidance strategy should be implemented instead. The core of this step lies in taking fundamental countermeasures after identifying interference from the same source, rather than simply switching links. When a deterioration in the regional electromagnetic environment is determined, it indicates that multiple heterogeneous communication links are simultaneously affected by external interference sources. At this point, the communication quality of all backup links has already deteriorated or is about to deteriorate. Performing link switching will not only fail to improve communication conditions but will also lead to communication interruptions, service loss, and wasted system resources due to frequent switching.

[0122] First, a link switching prohibition flag is set to prevent the automatic link switching module from triggering, thus avoiding a vicious cycle of ineffective switching. Simultaneously, a position avoidance strategy is immediately initiated. The basic principle of this strategy is to change the UAV's spatial position, moving it away from the interference source or altering the signal propagation path, thereby reducing interference intensity and restoring communication quality.

[0123] The specific implementation of the position avoidance strategy involves controlling the UAV to perform a vertical climb. The choice of a vertical upward climb direction is based on the following considerations: vertical climb can quickly increase the distance to ground interference sources while avoiding obstacles that may be encountered during horizontal movement; the electromagnetic environment in the vertical direction is generally better than in the horizontal direction because ground reflection and multipath effects weaken with increasing altitude. The climb rate is set to 2 m / s, which allows for rapid escape from the interference area without causing abrupt changes in aerodynamic loads that could affect flight stability due to excessively rapid climb.

[0124] During the climb, the maximum coherence value and the amplitude of the virtual vibration displacement sequence of each link are continuously monitored, with a sampling frequency maintained at 10Hz, to assess the changing trend of communication quality in real time. When the maximum coherence value drops below the adjusted threshold for determining common-source interference, and the amplitude of all links drops below the preset amplitude threshold, it is determined that the UAV has left the interference area. A stop-climb command is immediately sent, controlling the UAV to hover or switch to normal flight mode. At the same time, an alarm message is sent to the ground control station, including detailed data such as the timestamp of the interference occurrence, geographical coordinates, altitude information, duration, peak value of the maximum coherence value, and peak amplitude of each link, for subsequent analysis and interference source location.

[0125] This step fundamentally solves the problem of interference from the same source, rather than simply switching between affected links. Traditional methods automatically switch to backup links when a decline in communication quality is detected. However, in scenarios with interference from the same source, all links are affected. Switching merely shifts the problem from one link to another, failing to truly improve communication. In fact, the brief interruptions during the switching process can lead to data loss and task delays. By disabling link switching and implementing location avoidance, the impact of the interference source is eliminated from a spatial perspective, achieving a shift from passive response to proactive avoidance.

[0126] If the communication status monitoring concludes that a single link is faulty, a link switching strategy is implemented to switch communication services to a healthy link. This step takes targeted measures to address the independent fault of the link itself. When a single link fault is determined, it indicates that only a single link is experiencing a problem, while other links are functioning normally. In this case, link switching is the most effective response.

[0127] The execution process of the link switching strategy is as follows: First, faulty links and healthy links are identified. By traversing the amplitude data of each link, links with amplitudes exceeding a preset amplitude threshold are marked as faulty links, and links with amplitudes below the threshold are marked as healthy links. For example, if a 5G link with an amplitude of 1.4m is determined to be faulty, while a 2.4G link with an amplitude of 0.7m and a satellite link with an amplitude of 0.6m are determined to be healthy, then the 2.4G link and the satellite link are included in the switchable targets.

[0128] Next, the optimal handover target is selected from the healthy links. The selection criteria comprehensively consider the link's available bandwidth (BW), transmission delay (D), and signal quality (SNR), and a comprehensive score is calculated. ,in , , The weighting coefficients are 0.4, 0.3, and 0.3 respectively, and the link with the highest score is selected as the handover target. For example, if the 2.4G link score is 8.5 and the satellite link score is 6.2, then the 2.4G link is selected as the handover target.

[0129] Then, a "build-then-disconnect" handover mechanism is implemented. First, a new communication connection is established on the target link, completing the handshake and authentication process. Once the connection is confirmed to be stable, the connection to the faulty link is disconnected, and all communication services are migrated to the new link. This mechanism ensures communication continuity during the handover process, avoiding communication interruptions that might occur with the "disconnect-then-build" approach. The handover latency is controlled within 200ms, meeting the latency requirements for real-time UAV control.

[0130] After the switchover is complete, the faulty link undergoes self-checking and recovery attempts, including restarting the communication module, rescanning frequency points, and adjusting transmit power. If recovery is successful, it is reinstated into the available link pool. If multiple attempts fail to recover, it is marked as a permanent fault and an alarm is triggered. This step enables rapid fault isolation and seamless service migration, maximizing communication continuity.

[0131] In multi-link redundancy architectures, link switching is a fundamental means of addressing single-point failures. However, traditional switching strategies often rely on independent judgment of single-link metrics, lacking support from multi-link collaborative information, and are prone to ineffective switching under co-source interference scenarios. Through the aforementioned precise fault classification, this step can accurately identify scenarios suitable for link switching, avoiding ineffective switching under co-source interference and improving the targeting and effectiveness of the switching strategy.

[0132] If the communication status monitoring concludes that the communication status is normal, a maintenance strategy is executed to maintain the current communication configuration. This step ensures the stable operation of the system under normal conditions and avoids unnecessary intervention. When the communication status is determined to be normal, it indicates that the communication quality of all links is within the normal range and there are no abnormal situations that need to be addressed. At this time, the system should maintain the current link configuration, communication parameters, and operating mode unchanged and continue to perform normal monitoring tasks.

[0133] The maintenance strategy includes: keeping the current primary link unchanged and continuing to transmit control commands and task data through this link; keeping the backup link in standby mode, periodically sending heartbeat packets to detect link availability, but not carrying service traffic; keeping communication parameter configurations unchanged, including modulation method, coding rate, and transmit power; continuing to perform periodic communication status monitoring, maintaining a sampling frequency of 10Hz, continuously calculating indicators such as virtual vibration displacement sequence, spectrum function, and coherence function, and evaluating communication status in real time.

[0134] While seemingly simple, maintenance strategies are actually of great significance. In practical applications, communication parameters fluctuate randomly due to environmental factors. If the system is overly sensitive to these normal fluctuations, it will lead to frequent link switching, parameter adjustments, and alarm messages, increasing the system's computational burden and energy consumption, interfering with normal task execution, and reducing system availability. By clearly defining normal state determination and maintenance strategies, normal fluctuations are clearly distinguished from abnormal situations. Countermeasures are only triggered when an anomaly is confirmed, achieving a balance between monitoring sensitivity and system stability.

[0135] This step establishes a mechanism to ensure stable system operation, avoiding system instability caused by excessive intervention. In the design of the monitoring system, a precise balance must be struck between ensuring high sensitivity in detecting anomalies and avoiding overreaction to normal fluctuations. Through the aforementioned multi-dimensional index determination and dynamic threshold adjustment, this method can accurately distinguish between normal and abnormal states, maintaining stable system operation under normal conditions and triggering timely countermeasures under abnormal conditions, thus achieving intelligent state management.

[0136] This embodiment implements a classification control strategy based on communication status monitoring conclusions, establishing a complete closed-loop system from diagnosis to response. It effectively solves the problems of traditional methods' single and insufficiently targeted countermeasures, achieving precise policy implementation and efficient response. For scenarios with deteriorating regional electromagnetic environments, it prohibits invalid link switching and implements a location avoidance strategy, fundamentally solving the problem of co-source interference from a spatial perspective and preventing communication avalanche. For single-link failure scenarios, it executes rapid link switching to achieve fault isolation and service migration, ensuring communication continuity and demonstrating the advantages of a multi-link redundancy architecture. For normal states, it executes a maintenance strategy to keep the system running stably, avoiding excessive intervention and improving system availability. The entire strategy system adopts classification decision logic, executing different countermeasures according to different communication states, achieving precise strategy matching and efficient resource utilization. Compared to traditional single-response methods, this classification control strategy significantly improves the targeting and effectiveness of responses, avoiding invalid switching in co-source interference scenarios, achieving rapid recovery in single-link failure scenarios, and maintaining system stability in normal states. Its overall performance is significantly better than traditional methods, providing comprehensive technical support for the safe and reliable operation of UAVs in complex electromagnetic environments.

[0137] In one embodiment of this example, the position avoidance strategy includes the following steps: S810, controlling the UAV to perform a vertical climb operation; S820, continuously monitoring the maximum coherence value and the amplitude of the virtual vibration displacement sequence during the climb; S830, stopping the climb operation when the maximum coherence value drops below the same source interference determination threshold and the amplitude of the virtual vibration displacement sequence drops below the preset amplitude threshold.

[0138] In this embodiment, referring to FIG2, the climb height for the vertical climb operation is determined based on current environmental data and amplitude. The steps for determining the climb height for the vertical climb operation include: acquiring the current flight altitude and ground elevation of the UAV; calculating the average amplitude deviation value based on the amplitude of the virtual vibration displacement sequence of at least two heterogeneous communication links; calculating the interference intensity level based on the average amplitude deviation value, wherein the larger the average amplitude deviation value, the higher the interference intensity level; querying the corresponding reference climb height from a preset climb height mapping table according to the interference intensity level; acquiring the remaining battery power and task priority parameters of the UAV; calculating the climb height correction factor based on the remaining battery power and task priority parameters; and multiplying the reference climb height by the climb height correction factor to obtain the climb height for the vertical climb operation.

[0139] The system controls the drone to perform vertical climb maneuvers, actively avoiding interference from the same source by changing its spatial position. When the system determines that the regional electromagnetic environment has deteriorated, it indicates that the electromagnetic environment of the drone's current location is severely affected by external interference sources, and multiple communication links experience simultaneous quality degradation. In this case, moving away from the interference source through spatial maneuvering or changing the signal propagation path is the most effective response.

[0140] The choice of vertical climbing is based on the following technical considerations: First, the density of obstacles in the vertical direction is much lower than that in the horizontal direction. In urban environments, horizontal movement may encounter obstacles such as buildings and trees, while vertical ascent is usually an open space with higher safety. Second, electromagnetic interference sources are mostly located on or near the ground. As altitude increases, the strength of interference signals decreases according to the inverse square law. Vertical climbing can quickly increase the distance to the interference source. Third, ground reflection and multipath effects weaken with increasing altitude. The electromagnetic environment at high altitudes is generally better than at low altitudes, and signal propagation conditions are better.

[0141] Before performing a climb, the target climb altitude needs to be determined. This altitude is calculated by comprehensively considering factors such as interference intensity, UAV performance, and mission requirements. First, the current flight altitude of the UAV needs to be obtained. This data is obtained from barometric altimeter or GPS altitude information, and the unit is meters. Ground elevation is also acquired. This data is retrieved from a pre-installed terrain database or distributed via ground control stations to calculate the relative ground altitude, ensuring that the airspace restrictions are not exceeded after the ascent.

[0142] Next, the average amplitude deviation value is calculated to quantify the intensity of the current disturbance. The calculation formula is as follows: ,in For the number of links, For the first The amplitude of the virtual vibration displacement sequence of the link. The preset amplitude threshold is 1.0m. For example, if the amplitude of a 5G link is 1.3m, the amplitude of a 2.4G link is 1.5m, and the amplitude of a satellite link is 1.1m, then the average amplitude deviation value is... m. The larger the average amplitude deviation value, the more severe the interference to each link, and the greater the climb height is required to escape the interference area.

[0143] The interference intensity level is calculated based on the average amplitude deviation value, using a piecewise mapping method: when When m, the interference intensity level is 1 (slight); when When m, the level is 2 (moderate); when When m, the level is 3 (severe). For example, when At m, the interference intensity level is 2. The corresponding baseline climb height is retrieved from the preset climb height mapping table based on the interference intensity level. The mapping table was established based on a large amount of experimental data. Level 1 corresponds to a baseline climb height of 30m, Level 2 corresponds to 50m, and Level 3 corresponds to 80m. This step established a quantitative relationship between interference intensity and climb height, enabling evasive maneuvers to be precisely adjusted according to the actual interference situation. This avoids the problems of insufficient (inability to escape interference) or excessive (waste of energy and time) that may result from a fixed climb height.

[0144] Next, considering the actual constraints of the drone, the baseline climb altitude will be corrected. The remaining battery power of the drone will then be obtained. This data, provided by the battery management system, is expressed as a percentage, ranging from 0% to 100%. Task priority parameters are also retrieved. This parameter is set by the mission planning module and can be 1 (low priority), 2 (medium priority), or 3 (high priority). High-priority missions (such as emergency rescue and important reconnaissance) have higher requirements for communication reliability and can tolerate greater energy consumption; low-priority missions (such as routine inspections) need to balance communication quality and endurance.

[0145] Calculate the climb height correction factor based on remaining battery power and task priority. The first term of the formula This indicates battery level correction; the value is 1 when the battery is 100% and 0.5 when the battery is 0%, ensuring that a minimum level of protection is maintained even when the battery is low. The second item... This indicates priority adjustment. When priority is 3, this value is 1; when priority is 1, it is approximately 0.87. Higher priority tasks can perform larger ramp-ups. For example, when the remaining battery is 60% and the task priority is 2, the adjustment factor is... .

[0146] Multiply the baseline climb height by the correction factor to obtain the final climb height. For example, with a baseline climb altitude of 50m and a correction factor of 0.744, the target climb altitude is 37.2m. This dynamic adjustment mechanism achieves adaptive matching between the evasion strategy and the UAV's state, balancing energy efficiency and mission requirements while ensuring escape from interference, and avoiding resource waste or safety risks that might result from a fixed strategy. After determining the target climb altitude, a climb command is sent to the flight control system, containing the target altitude and climb rate (2m / s). Upon receiving the command, the flight control system adjusts the motor speed, controlling the UAV to ascend vertically at a constant rate while maintaining its horizontal position and heading, ensuring the stability and controllability of the climb process.

[0147] During the climb, the maximum coherence value and the amplitude of the virtual vibration displacement sequence are continuously monitored to assess the changing trend of communication quality in real time, providing a basis for determining whether the area has escaped the interference zone. The monitoring process maintains the same sampling frequency of 10Hz as during normal flight, that is, the physical layer parameters of all links are collected every 100 milliseconds to calculate indicators such as virtual vibration displacement sequence, spectral function, and coherence function.

[0148] During the ascent, as the drone's altitude increases and the distance to ground-based interference sources increases, the interference signal strength gradually weakens, and the communication quality of each link gradually improves. This improvement is reflected in the following monitoring indicators: the real-time signal-to-noise ratio of each link gradually increases, transmission delay gradually decreases, and available bandwidth gradually increases; the excitation force and amplitude of the corresponding virtual vibration displacement sequence gradually decrease; the peak value at the interference frequency in the spectral function gradually weakens; the cross-power spectral density gradually decreases, and the coherence function value gradually declines. By continuously monitoring these changes, the avoidance effect can be assessed in real time, and it can be determined whether the current altitude has escaped the interference zone.

[0149] The specific monitoring process is as follows: During each sampling period, physical layer parameters of all links are collected synchronously. The virtual vibration displacement of each link is calculated using the aforementioned method, and the displacement value at the current moment is extracted as the instantaneous amplitude. Due to the rapid changes in communication status during the climb process, the 10-second time window is no longer used for frequency domain analysis. Instead, a sliding window method is adopted, with the window length shortened to 5 seconds (50 sampling points). The window data is updated once per sampling period, and the spectral function and coherence function of the data within the window are calculated to obtain the real-time maximum coherence value. Simultaneously, the maximum amplitude of each link within the current 5-second window is recorded.

[0150] The maximum coherence value monitored in real time The adjusted threshold for determining co-source interference Compare the maximum amplitude of each link. With preset amplitude threshold Comparison. When And all links Upon reaching the designated point, the system determines that the system has escaped the interference zone, triggering a stop-climb condition. To avoid misjudgments caused by instantaneous fluctuations, a stability verification mechanism is introduced: only when the above conditions are met consecutively for three sampling periods (0.3 seconds) is it confirmed that the system has escaped interference and triggers the stop-climb operation. For example, when climbing to a height of 40m relative to the starting position, if the maximum coherence value decreases from 0.88 to 0.68, the 5G link amplitude decreases from 1.3m to 0.9m, the 2.4G link amplitude decreases from 1.5m to 0.8m, and the satellite link amplitude decreases from 1.1m to 0.7m, the escape condition is met, and the system remains stable for three consecutive periods, confirming escape from interference.

[0151] This step enables real-time evaluation and dynamic feedback of the evasion effect, allowing the evasion process to adaptively adjust based on the actual results. Compared to methods that preset a fixed climb height, this real-time monitoring mechanism can dynamically determine the stopping point based on the actual interference distribution and attenuation patterns. This avoids both evasion failure due to insufficient climb and energy waste due to excessive climb, achieving an optimal balance between evasion effectiveness and resource consumption.

[0152] When the maximum coherence value decreases below the threshold for determining common-source interference, and the amplitude of the virtual vibration displacement sequence decreases below the preset amplitude threshold, the climb operation stops, completing the avoidance process. This step executes the termination judgment and subsequent processing of the avoidance strategy. Once the real-time monitoring system confirms that the conditions for escaping interference have been met for three consecutive sampling cycles, it immediately sends a stop-climb command to the flight control system. Upon receiving the command, the flight control system gradually reduces its vertical speed, decreasing the climb rate from 2 m / s to 0 within 2 seconds, allowing the UAV to smoothly transition to a hovering state and avoid attitude disturbances caused by a sudden stop.

[0153] The hovering state is maintained for 5 seconds, during which communication quality is continuously monitored to verify the stability of escaping interference. If the communication quality remains stable within 5 seconds and none of the indicators exceed the threshold again, the avoidance is confirmed as successful, the link switching prohibition flag is removed, and normal communication monitoring and link management mode is restored. If the communication quality deteriorates again within 5 seconds and the indicators exceed the threshold again, it indicates that the current altitude is still within the interference range, and further ascent is required. In this case, the target ascent altitude is recalculated, increased by 20m from the current altitude, and the ascent operation continues until it is confirmed that the interference has been escaped.

[0154] After successful evasion, a detailed alarm report is sent to the ground control station, including: the timestamp and geographic coordinates of the interference occurrence, recording the approximate location of the interference source; the duration of the interference, the total time from the detection of regional electromagnetic environment deterioration to confirmation of escape from the interference; interference characteristic parameters, including the peak value of the maximum coherence value, the peak amplitude of each link, and the frequency distribution characteristics of the coherence function, for interference source identification and classification; evasion process parameters, including the initial altitude, final altitude, climb distance, and power consumption, for evaluating the effectiveness and cost of the evasion strategy; and the current UAV status, including position, altitude, remaining power, and communication link status, for the ground station to assess the ability to perform subsequent tasks.

[0155] After receiving alarm reports, the ground control station can perform interference source location analysis. If multiple drones report similar interference characteristics from different locations, the location of the interference source can be determined through triangulation or multi-point intersection methods, providing a basis for subsequent interference source investigation and elimination. Simultaneously, the ground station can decide whether to adjust the mission plan based on the drone's current status and remaining tasks, such as changing the flight path to avoid the interference area, reducing mission priority to save power, or recalling drones to replace batteries. This step establishes a complete closed loop in the avoidance process, forming an automated processing flow from triggering avoidance, executing avoidance, monitoring effects to confirming completion. Furthermore, detailed alarm reports and data records provide data support for interference source analysis, strategy optimization, and system improvement, achieving an improvement in capabilities from single-response to continuous improvement.

[0156] This embodiment implements a complete position avoidance strategy, fundamentally solving the problem of co-source interference. It eliminates the influence of the interference source through spatial maneuvering, avoiding the ineffectiveness of traditional link switching methods in co-source interference scenarios. By comprehensively considering interference intensity, UAV status, and mission requirements, and dynamically calculating the target climb altitude, the avoidance strategy is made precise and personalized, avoiding the limitations of fixed strategies. The introduction of the average amplitude deviation value accurately quantifies the interference intensity, and the introduction of the climb altitude correction factor achieves adaptive matching with the UAV status, enabling the avoidance action to effectively escape interference while also considering energy efficiency and mission requirements. Continuous monitoring of communication quality indicators enables real-time evaluation and dynamic feedback of the avoidance effect, allowing the avoidance process to adaptively adjust the stopping point based on the actual effect, avoiding insufficient or excessive climb, and achieving an optimal balance between avoidance effect and resource consumption. The application of the sliding window method improves the real-time performance of monitoring, and the stability verification mechanism avoids misjudgments caused by instantaneous fluctuations, ensuring the accuracy of the stopping decision. A smooth stopping process and stability verification ensure reliable completion of the avoidance, and detailed alarm reports provide data support for subsequent analysis and improvement. The entire position avoidance strategy embodies proactive defense, taking fundamental countermeasures at the early stages of detecting interference from the same source, thus preventing system collapse caused by continuous deterioration of communication quality. Compared to traditional passive response methods, this strategy advances the response time from after communication interruption to when communication quality fluctuates slightly, and upgrades the response method from link switching to spatial avoidance, significantly improving the effectiveness and proactivity of the response, and providing strong technical support for the safe and reliable operation of UAVs in complex electromagnetic environments.

[0157] This application also provides a drone, including: a memory configured to store instructions; and an onboard processor configured to retrieve the instructions from the memory and, when executing the instructions, to implement the above-described multi-link-based drone flight communication status monitoring method.

[0158] Referring to Figure 3, this application embodiment also provides a multi-link-based UAV flight communication status monitoring system, including: a UAV equipped with at least two heterogeneous communication links; and a ground control station that communicates with the UAV through at least two heterogeneous communication links.

[0159] In this embodiment, as shown in Figure 3, the upper part of the figure represents the UAV, which is equipped with a communication status monitoring module and functional components for real-time monitoring of the status of each link. This module can collect key parameters such as signal strength, packet loss rate, and latency. The UAV establishes communication connections with the ground control station through three heterogeneous link interfaces: Heterogeneous link interface 1 connects to the 4G / 5G mobile communication link of link A, which features high bandwidth and low latency; heterogeneous link interface 2 connects to the satellite communication link of link B, providing wide coverage and high reliability communication; heterogeneous link interface 2 can also connect to the air-to-ground data transmission link of link C, using a dedicated frequency band to achieve point-to-point communication. Each link can transmit flight status data and communication quality information bidirectionally, while simultaneously receiving control commands and link configuration information from the ground. The lower part of the figure represents the ground control station, which includes heterogeneous link access points corresponding to the UAV: ​​Heterogeneous link access 1 connects to the 4G / 5G network through a public network base station or core network, and heterogeneous link access 2 establishes communication connections with the UAV through a satellite ground station or data transmission radio. The ground control station is equipped with a communication status processing and display platform. This platform has functions such as multi-link status fusion analysis, fault early warning, and intelligent switching strategies, and can visualize the operational status of each link in real time. The UAV is equipped with cellular network communication links, private frequency band data transmission radio communication links, and satellite communication links. It acquires communication status data of each link in real time through the communication module in the onboard processor and works in collaboration with the ground control station to achieve comprehensive monitoring and intelligent management of the multi-link communication status. This ensures the communication reliability and flight safety of the UAV in complex environments and provides a complete hardware foundation and data source for coherence analysis and common source interference identification.

[0160] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0161] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, as well as combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in one or more blocks of the flowchart illustrations and / or one or more blocks of the block diagrams.

[0162] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means that implement the functions specified in one or more flowcharts and / or one or more block diagrams.

[0163] These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer-implemented process, such that the instructions, which execute on the computer or other programmable apparatus, provide steps for implementing the functions specified in one or more flowcharts and / or one or more block diagrams.

[0164] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0165] Memory may include non-persistent memory in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.

[0166] Computer-readable media includes both permanent and non-permanent, removable and non-removable media that can store information using any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0167] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.

[0168] The above are merely embodiments of this application and are not intended to limit the scope of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of the claims of this application.

Claims

1. A method for monitoring the flight communication status of a UAV based on multiple links, characterized in that, This method, applied to unmanned aerial vehicles (UAVs) equipped with at least two heterogeneous communication links, includes: simultaneously acquiring physical layer parameters of at least two heterogeneous communication links; mapping the physical layer parameters to virtual vibration displacement sequences for each heterogeneous communication link; extracting the amplitude of the virtual vibration displacement sequences within a preset time window and performing frequency domain transformation on the virtual vibration displacement sequences of the at least two heterogeneous communication links to obtain the corresponding spectral functions; calculating the cross-power spectral density between the spectral functions of different heterogeneous communication links and determining the coherence function and the maximum coherence value of the coherence function based on the cross-power spectral density; acquiring the flight attitude data of the UAV and dynamically adjusting a preset threshold for determining common-source interference based on the flight attitude data; generating a communication status monitoring conclusion based on the maximum coherence value, the adjusted threshold for determining common-source interference, and the amplitude; and executing a corresponding link control strategy based on the communication status monitoring conclusion, wherein the link control strategy includes a position avoidance strategy, a link switching strategy, and a maintenance strategy.

2. The method according to claim 1, characterized in that, The physical layer parameters include real-time signal-to-noise ratio, transmission delay, and available bandwidth. For each heterogeneous communication link, the physical layer parameters are mapped to a virtual vibration displacement sequence, including: determining virtual mass parameters for each heterogeneous communication link; mapping the available bandwidth of the heterogeneous communication link to a stiffness coefficient and the transmission delay to a damping coefficient, where the stiffness coefficient is positively correlated with the available bandwidth and the damping coefficient is positively correlated with the transmission delay; using the reciprocal of the real-time signal-to-noise ratio of the heterogeneous communication link as the excitation force function; constructing a second-order differential equation, which includes the virtual mass parameter, stiffness coefficient, damping coefficient, and excitation force function; and iteratively solving the second-order differential equation to generate the virtual vibration displacement sequence of the heterogeneous communication link.

3. The method according to claim 1, characterized in that, The heterogeneous communication links include a first heterogeneous communication link and a second heterogeneous communication link. The cross-power spectral density between the spectral functions of the different heterogeneous communication links is calculated, and the coherence function and its maximum coherence value are determined based on the cross-power spectral density. This includes: performing windowed Fourier transforms on the virtual vibration displacement sequences of the first and second heterogeneous communication links respectively to obtain a first spectral function and a second spectral function; calculating the cross-power spectral density based on the first and second spectral functions; calculating the first auto-power spectral density of the first spectral function and the second auto-power spectral density of the second spectral function respectively; calculating the coherence function based on the cross-power spectral density, the first auto-power spectral density, and the second auto-power spectral density; and iterating through the coherence values ​​of the coherence function at different frequency points to extract the maximum coherence value.

4. The method according to claim 3, characterized in that, The coherence function is calculated based on the cross-power spectral density, the first self-power spectral density, and the second self-power spectral density, including: calculating the square of the modulus of the cross-power spectral density; calculating the product of the first self-power spectral density and the second self-power spectral density; and dividing the square of the modulus of the cross-power spectral density by the product to obtain the coherence function.

5. The method according to claim 1, characterized in that, Flight attitude data includes three-axis acceleration and three-axis angular velocity. The preset threshold for determining common-source interference is dynamically adjusted based on this data, including: calculating the UAV's total body acceleration based on the three-axis acceleration; calculating the maneuver intensity factor based on the total body acceleration; calculating the UAV's angular velocity magnitude based on the three-axis angular velocity; calculating the attitude change rate based on the angular velocity magnitude; calculating the comprehensive maneuver index based on the maneuver intensity factor and the attitude change rate; calculating the threshold adjustment coefficient based on the comprehensive maneuver index, where a higher comprehensive maneuver index results in a larger threshold adjustment coefficient; and multiplying the preset threshold for determining common-source interference by the threshold adjustment coefficient to obtain the adjusted threshold for determining common-source interference.

6. The method according to claim 1, characterized in that, Based on the maximum coherence value, the adjusted same-source interference judgment threshold, and the amplitude, a communication status monitoring conclusion is generated, including: if the maximum coherence value exceeds the adjusted same-source interference judgment threshold, and the amplitude of the virtual vibration displacement sequence of at least two heterogeneous communication links exceeds the preset amplitude threshold, the communication status monitoring conclusion is determined to be a deterioration of the regional electromagnetic environment; if the maximum coherence value does not exceed the same-source interference judgment threshold, and the amplitude of the virtual vibration displacement sequence of a single heterogeneous communication link exceeds the preset amplitude threshold, the communication status monitoring conclusion is determined to be a single-link failure; if the maximum coherence value does not exceed the same-source interference judgment threshold, and the amplitude of the virtual vibration displacement sequence of all heterogeneous communication links does not exceed the preset amplitude threshold, the communication status monitoring conclusion is determined to be normal.

7. The method according to claim 6, characterized in that, Based on the communication status monitoring conclusions, the corresponding link control strategies are executed, including: if the communication status monitoring conclusions indicate a deterioration in the regional electromagnetic environment, the link switching strategy is prohibited, and a location avoidance strategy is executed; if the communication status monitoring conclusions indicate a single link failure, the link switching strategy is executed to switch the communication service to a healthy link; if the communication status monitoring conclusions indicate a normal communication status, a maintenance strategy is executed to maintain the current communication configuration.

8. The method according to claim 7, characterized in that, The position avoidance strategy includes: controlling the drone to perform a vertical climb operation; continuously monitoring the maximum coherence value and the amplitude of the virtual vibration displacement sequence during the climb; and stopping the climb operation when the maximum coherence value drops below the threshold for determining common source interference and the amplitude of the virtual vibration displacement sequence drops below the preset amplitude threshold.

9. A drone, characterized in that, include: The memory is configured to store instructions; And an onboard processor configured to retrieve the instructions from the memory and, when executing the instructions, to implement the multi-link-based UAV flight communication status monitoring method according to any one of claims 1 to 8.

10. A multi-link-based UAV flight communication status monitoring system, characterized in that, include: The UAV according to claim 9 is equipped with at least two heterogeneous communication links; The ground control station communicates with the UAV through at least two heterogeneous communication links.

Images