A frequency adaptive switching system and method for unmanned aerial vehicles (UAVs)

By installing a spectrum sensing and decision-making module on the UAV's onboard terminal, interference levels can be collected and predicted in real time, enabling seamless frequency switching. This solves the problem of communication interruption for UAVs in complex electromagnetic environments and improves the continuity and safety of operations.

CN123124031APending Publication Date: 2026-11-17STATE GRID JIANGSU ELECTRIC POWER CO LTD TAIZHOU POWER SUPPLY BRANCH +2
View PDF 1 Cites 0 Cited by

Patent Information

Application Number
CN202611185992.9
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-08-06
Publication Date
2026-11-17

AI Technical Summary

Technical Problem

Existing UAV communication systems cannot achieve efficient and seamless frequency switching in the face of complex electromagnetic interference environments, resulting in communication interruptions and reduced operational continuity, and failing to fully utilize the onboard computing capabilities of industry-grade UAVs.

Method used

By installing a spectrum sensing module and an airborne decision module on the UAV, and using a periodic spectrum scanning strategy and electromagnetic environment map construction, interference levels and channel quality parameters are collected in real time. Combined with throughput prediction and hot backup switching functions, adaptive switching of frequency points is achieved, eliminating processing delays at the ground control end.

Benefits of technology

It improves the measurement accuracy and response speed of frequency selection, ensures seamless communication switching, reduces the computational load on the ground control terminal, and enhances the continuity and safety of operations.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123124031A_ABST
    Figure CN123124031A_ABST
Patent Text Reader

Abstract

This invention discloses a frequency adaptive switching system and method for unmanned aerial vehicles (UAVs). The system includes a ground control terminal and an airborne terminal. The ground control terminal monitors the channel quality parameters of the current communication link in real time, judges the channel quality parameters based on a pre-configured list of backup frequencies, and generates a switching trigger signal. It also constructs an electromagnetic environment map and records the throughput correction coefficients for each geographical location. The airborne terminal uses a periodic spectrum scanning strategy to collect the interference levels of each candidate sub-band in real time. Based on the signal power mapping relationship, and combined with the interference levels and throughput correction coefficients of each candidate sub-band, it predicts the achievable throughput of each candidate sub-band. Based on the achievable throughput of each candidate sub-band, it determines the target frequency and executes the frequency switching operation by combining the switching trigger signal and a collaborative decision-making strategy. This system utilizes industry-grade UAV onboard computing capabilities to achieve distributed decision-making, eliminates ground-end processing latency, and enables seamless frequency switching for UAVs.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention belongs to the technical field of unmanned aerial vehicle (UAV) communication, specifically relating to a frequency adaptive switching system and method for UAVs. Background Technology

[0002] With the rapid development of drone technology, industrial-grade drones are increasingly being used in fields such as power line inspection, oil and gas pipeline inspection, security monitoring, emergency rescue, and industrial inspection. Unlike consumer-grade drones, industrial-grade drones place higher demands on the reliability, real-time performance, and anti-interference capabilities of communication links. Their application scenarios typically involve strong electromagnetic interference and complex flight environments, requiring high bandwidth and high reliability.

[0003] Patent application CN113301654A discloses a UAV communication frequency band allocation system and method based on a semi-supervised learning algorithm. The system includes a ground control center and an airborne control terminal. The method involves the ground control center first performing a spectrum scan on the pre-used frequency band and determining whether interference exists on that band. If interference is present, the system switches to the next frequency band for spectrum scanning. This method ensures that the UAV can communicate reliably with the ground control center before actual flight. Then, during normal operation, the UAV's airborne control terminal continuously monitors the current communication frequency band information and operational status, and predicts the idle status of other frequency bands. When the current communication frequency band is interfered with, the system can quickly switch bands, avoiding unnecessary interference to the UAV's normal telemetry and data transmission, thereby improving the UAV's operating efficiency and anti-interference capability.

[0004] Including the aforementioned technologies, the switching decisions are mainly based on the received signal power, without considering interference levels for throughput prediction. They also fail to achieve seamless switching of hot backup frequency points and do not fully utilize the onboard computing power of industry-grade UAV platforms to achieve distributed decision-making. How to leverage the onboard computing power of industry-grade UAVs to achieve distributed decision-making, eliminate ground-end processing delays, and simultaneously achieve seamless switching of UAV frequencies is a problem that needs to be solved. Summary of the Invention

[0005] To address the shortcomings of the existing technology, this invention provides a frequency adaptive switching system and method for unmanned aerial vehicles (UAVs). The system includes a ground control terminal and an airborne terminal. The ground control terminal is used to monitor the channel quality parameters of the current communication link in real time, determine the channel quality parameters based on a pre-configured list of backup frequencies, and generate a switching trigger signal. It also constructs an electromagnetic environment map and records the throughput correction coefficients for each geographical location. The airborne terminal is used to collect the interference levels of each candidate sub-band in real time using a periodic spectrum scanning strategy, wherein the periodic spectrum scanning strategy is determined by judging the channel quality parameters of the current communication link. It also predicts the achievable throughput of each candidate sub-band based on the signal power mapping relationship, combined with the interference levels and throughput correction coefficients of each candidate sub-band. Finally, it determines the target frequency point based on the achievable throughput of each candidate sub-band and performs the frequency switching operation in conjunction with the switching trigger signal and a collaborative decision-making strategy.

[0006] By installing a spectrum sensing module on the airborne terminal, the UAV periodically calls the spectrum scanning interface provided by the industry-grade UAV SDK to obtain the interference levels of each candidate sub-band within a specified frequency band. Simultaneously, to balance the measurement accuracy and system overhead of the UAV's airborne terminal, a dynamic periodic spectrum scanning strategy is adopted. This strategy enables adaptive adjustment of the scanning cycle, allowing the system to automatically adjust the strategy according to environmental changes and maintain optimal performance at all times. Compared to ground control measurements or static electromagnetic maps, this improves measurement accuracy and provides a more accurate basis for frequency selection. Utilizing the hot backup switching function of the industry-grade UAV image transmission system, seamless switching is achieved, with uninterrupted image transmission and smooth remote control during switching, significantly improving operational continuity and safety. By updating the backup frequency list after each spectrum scan, the timeliness of information on each candidate frequency in the backup frequency list is ensured.

[0007] By installing an airborne decision-making module, the processing delay at the ground control end can be eliminated, allowing decisions to be made on-board and unaffected by communication delays at the ground control end. Simultaneously, response speed is improved; the system can execute decisions directly upon receiving the handover trigger signal, without waiting for a response from the ground control end. Furthermore, even in the event of a communication interruption with the ground control end, the airborne end can still complete the handover decision, reducing the computational load on the ground control end.

[0008] In a first aspect, the present invention provides a frequency adaptive switching system for unmanned aerial vehicles (UAVs), comprising a ground control terminal and an airborne terminal; The ground control terminal is used to monitor the channel quality parameters of the current communication link in real time. Combined with the list of backup frequency points pre-configured for hot backup frequency points, it judges the channel quality parameters and generates a handover trigger signal. In addition, construct an electromagnetic environment map and record the throughput correction coefficient for each geographical location; The airborne terminal is used to collect the interference level of each candidate sub-band in real time through a periodic spectrum scanning strategy. The periodic spectrum scanning strategy is determined by judging the channel quality parameters of the current communication link. Furthermore, based on the signal power mapping relationship, and combined with the interference level and throughput correction coefficient of each candidate sub-band, the achievable throughput of each candidate sub-band is predicted. Furthermore, the target frequency is determined by the available throughput of each candidate subband, and the frequency switching operation is performed by combining the switching trigger signal and the collaborative decision-making strategy.

[0009] Furthermore, the periodic spectrum scanning strategy is determined by judging the channel quality parameters of the current communication link, and is determined through the following steps: Construct the correlation between channel quality parameters and scanning period for each communication link to form a spectrum scanning period library; Based on the channel quality parameters of the current communication link, the corresponding scanning period is matched from the spectrum scanning period library to determine the periodic spectrum scanning strategy.

[0010] Furthermore, the electromagnetic environment map is determined through the following steps: Read spectrum measurement data from multiple drones. The spectrum measurement data includes location data, frequency band information, sub-band information, interference level, signal-to-noise ratio, and timestamp. An interpolation algorithm is used to fuse spectrum measurement data from multiple UAVs to construct an electromagnetic environment map.

[0011] Furthermore, an interpolation algorithm is used to fuse the spectrum measurement data of multiple UAVs to construct an electromagnetic environment map, specifically including: The position data of each drone is obtained based on the spectrum measurement data of multiple drones; Analyze the location data of each UAV and the distance between the location of the point to be interpolated in the electromagnetic environment map, and combine the Gaussian kernel smoothing coefficient to give the corresponding measurement weight for each UAV. Based on the measurement weights of each UAV, the spectrum measurement data are fused to obtain the spectrum data of each interpolation point, thus forming an electromagnetic environment map.

[0012] Furthermore, the spectral data of each interpolation point is specifically represented as follows:

[0013] Among them, I map (p) represents the spectral data of the point to be interpolated corresponding to location data p in the electromagnetic environment map, where K is the total number of sampling points for each UAV. k Here, σ represents the location data corresponding to sampling point k, and I represents the Gaussian kernel smoothing coefficient. k For sampling point k at p kThe spectral measurement data at the location is given by exp(), which is an exponential function.

[0014] Furthermore, based on the pre-configured list of backup frequency points for hot backup, channel quality parameters are assessed, and a handover trigger signal is generated, specifically including: Based on a preset hysteresis time window, the channel quality parameters of the current communication link are determined: If the channel quality parameters of the current communication link are lower than the service threshold, a first handover signal is generated. Based on the first switching signal, the first backup frequency point is determined from the list of backup frequency points pre-configured for the hot backup frequency point, and the throughput of the first backup frequency point is judged to give a switching trigger signal.

[0015] Furthermore, the list of pre-configured spare frequency points for hot backup frequency points is determined through the following steps: The backup frequency list maintenance unit is used to maintain the first backup frequency, store the frequency identifier and score, and the list structure includes the current working frequency, the first backup frequency and the score of each frequency. The dynamic update unit is used to periodically evaluate all candidate frequency points and set the candidate frequency point with the highest predicted throughput as the first backup frequency point. The pre-switching unit is used to actively call the SDK interface to pre-configure the first backup frequency point when the backup frequency point score exceeds the current working frequency point, but without triggering an actual switch. The seamless switching execution unit is used to call the SDK hot backup switching interface when a switch is triggered to perform a seamless switch.

[0016] Furthermore, all candidate frequencies are periodically evaluated, and the candidate frequency with the highest predicted throughput is designated as the first backup frequency. Specifically, this includes: Obtain all current candidate frequency points and form a candidate frequency point list; Based on a preset update cycle, and combining an electromagnetic environment map and a throughput prediction function, the predicted throughput of each candidate frequency point in the candidate frequency point list is calculated. Based on the predicted throughput of each candidate frequency point, sort the candidate frequency points in the candidate frequency point list and update the order of each candidate frequency point in the candidate frequency point list; Set the candidate frequency with the highest predicted throughput from the candidate frequency list as the first backup frequency; Based on the update cycle, the predicted throughput of each candidate frequency point is updated in real time, and the first backup frequency point is dynamically replaced.

[0017] Furthermore, the signal power mapping relationship is determined through the following steps: Read the actual received power at different frequencies in each frequency band and the signal strength indication value of the drone at different frequencies from the cloud; By establishing the correspondence between the actual received power and the signal strength indication values, data pairs are formed, and a set of collected data points is constructed. Based on the collected data points, the actual received power and signal strength indication values ​​in different frequency bands are fitted to obtain the mapping relationship parameters for different frequency bands; By combining the drone's model identifier, a mapping relationship between the model identifier and different frequency bands is constructed to form a signal power mapping relationship.

[0018] Furthermore, based on the signal power mapping relationship, and combined with the interference level and throughput correction coefficient of each candidate sub-band, the achievable throughput of each candidate sub-band is predicted, specifically including: By using the signal power mapping relationship and combining the signal strength indication values ​​of each frequency band, the actual received power of each frequency band can be determined. Based on the actual received power of each frequency band and the interference level of each candidate sub-band, the effective signal-to-noise ratio of each candidate sub-band is calculated. Based on the effective signal-to-noise ratio of each candidate subband, match the corresponding throughput prediction function; Substitute the effective signal-to-noise ratio of each candidate subband into the corresponding throughput prediction function to obtain the first predicted throughput value of each candidate subband. Based on the current location of the drone, determine the throughput correction factor for the current area; By using a throughput correction factor, the first predicted value of throughput is corrected, and the achievable throughput of each candidate subband is given.

[0019] Furthermore, the throughput prediction function is specifically expressed as:

[0020] Among them, T base SNR is the initial throughput prediction value. eff For the effective signal-to-noise ratio, β1 is the first prediction coefficient, T1 is the first prediction constant, γ1 is the first prediction threshold, β2 is the second prediction coefficient, T2 is the second prediction constant, γ2 is the second prediction threshold, β3 is the third prediction coefficient, T3 is the third prediction constant, γ3 is the third prediction threshold, T4 is the fourth prediction constant, γ4 is the fourth prediction threshold, and u(x) is the unit step function.

[0021] Furthermore, the frequency switching operation is performed by combining the switching trigger signal and the collaborative decision-making strategy, specifically including: Acquire the target frequency point on the airborne end and the first backup frequency point for the switching trigger signal on the ground control end; If the first backup frequency point and the target frequency point are the same, then execute the switching trigger signal to complete the frequency point switching; If the first backup frequency and the target frequency are different, the communication quality is determined based on the channel quality parameters of the current communication link, and the first backup frequency or the target frequency is selected to perform frequency switching.

[0022] Secondly, the present invention also provides a frequency adaptive switching method for a drone, applied to a frequency adaptive switching system for a drone as described in any of the above claims, comprising: The channel quality parameters of the current communication link are monitored in real time. Combined with the list of backup frequency points pre-configured for hot backup frequency points, the channel quality parameters are judged and a handover trigger signal is generated. The interference levels of each candidate subband are collected in real time using a periodic spectrum scanning strategy. Based on the signal power mapping relationship, and combined with the interference level and throughput correction coefficient of each candidate sub-band, the achievable throughput of each candidate sub-band is predicted. The target frequency is determined by the available throughput of each candidate subband, and the frequency switching operation is performed in conjunction with the switching trigger signal.

[0023] The present invention provides a frequency adaptive switching system and method for unmanned aerial vehicles (UAVs), which has at least the following beneficial effects: (1) By setting up a spectrum sensing module on the airborne end, the spectrum scanning interface provided by the industry-grade UAV SDK is periodically called in the UAV to obtain the interference level of each candidate sub-band within the specified frequency band. At the same time, in order to balance the measurement accuracy and system overhead of the UAV airborne end, a dynamic periodic spectrum scanning strategy is adopted. The scanning period is adaptively adjusted through the periodic spectrum scanning strategy, and the system can automatically adjust the strategy according to environmental changes to always maintain optimal performance. Compared with ground control end measurement or static electromagnetic map, the measurement accuracy is improved, providing a more accurate decision basis for frequency point selection.

[0024] (2) Data sharing and multi-machine collaborative electromagnetic environment map construction are achieved through the cloud. Multiple drones can share data and interference information, and accurate predictions can be obtained on the first entry into the area, reducing the cost of repeated calibration.

[0025] (3) By utilizing the hot backup switching function of the industry-grade UAV image transmission system, seamless switching is achieved, ensuring uninterrupted image transmission and smooth remote control during the switching process, significantly improving operational continuity and safety. By updating the list of backup frequency points after each spectrum scan, the timeliness of information on each candidate frequency point in the backup frequency point list is guaranteed.

[0026] (4) By setting up an airborne decision-making module, the processing delay of the ground control terminal can be eliminated, and the decision can be completed on the airborne terminal without being affected by the communication delay of the ground control terminal. At the same time, the response speed can be improved, and the handover trigger signal can be received directly without waiting for the response of the ground control terminal. Furthermore, the airborne terminal can still complete the handover decision even if the communication with the ground control terminal is interrupted, reducing the computational load of the ground control terminal. Attached Figure Description

[0027] Figure 1 This is a structural framework diagram of a frequency adaptive switching system for an unmanned aerial vehicle (UAV) provided in an embodiment of the present invention. Figure 2 An example diagram of the periodic spectrum scanning strategy provided in the embodiments of the present invention; Figure 3 An example diagram illustrating the relationship between the obtainable throughput and the throughput correction coefficient provided in an embodiment of the present invention; Figure 4 Example diagrams illustrating system performance effects provided in embodiments of the present invention; Figure 5 A flowchart of a frequency adaptive switching method for a drone provided in an embodiment of the present invention. Detailed Implementation

[0028] To better understand the above technical solutions, a detailed description of the solutions will be provided below in conjunction with the accompanying drawings and specific embodiments. Obviously, the described embodiments are merely some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0029] The terminology used in the embodiments of this invention is for the purpose of describing particular embodiments only and is not intended to limit the invention. The singular forms “a,” “the,” and “the” as used in the embodiments of this invention and the appended claims are also intended to include the plural forms, and “multiple” generally includes at least two unless the context clearly indicates otherwise.

[0030] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that an article or device that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such an article or device. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the article or device that includes said element.

[0031] The frequency adaptive switching system for drones provided in this invention is applied to industrial-grade drones. Taking the DJI M350 RTK industrial-grade drone as an example, it is used for power inspection tasks in substations. The drone is a DJI M350 RTK, equipped with the OcuSync Enterprise image transmission system. The drone is also connected to an external DJI Manifold 2 onboard computer for onboard computing. The ground control terminal is the DJI RC Pro Enterprise Edition, with candidate frequency bands including 1.8GHz, 2.4GHz, and 5.8GHz, and a subband bandwidth of 20MHz. The service requirement is 4K high-definition video transmission, with a stable throughput of ≥20Mbps.

[0032] like Figure 1 As shown, this embodiment of the invention provides a frequency adaptive switching system for unmanned aerial vehicles (UAVs), including a ground control terminal and an airborne terminal. The ground control terminal includes a communication quality monitoring module, an electromagnetic map construction module, and a switching trigger module. The communication quality monitoring module monitors the channel quality parameters of the current communication link in real time. The switching trigger module, combined with a pre-configured list of backup frequencies, judges the channel quality parameters and generates a switching trigger signal. The electromagnetic map construction module constructs an electromagnetic environment map and records the throughput correction coefficients for each geographical location. The airborne terminal includes a spectrum sensing module, a throughput prediction module, and an airborne decision module. The spectrum sensing module collects the interference levels of each candidate sub-band in real time using a periodic spectrum scanning strategy, where the periodic spectrum scanning strategy is determined by judging the channel quality parameters of the current communication link. The throughput prediction module predicts the achievable throughput of each candidate sub-band based on the signal power mapping relationship, combined with the interference levels and throughput correction coefficients of each candidate sub-band. The airborne decision module determines the target frequency based on the achievable throughput of each candidate sub-band and executes the frequency switching operation by combining the switching trigger signal and a collaborative decision-making strategy.

[0033] It's important to understand that candidate subbands refer to the continuous frequency range supported by the device, retained after compliance and interference screening, such as 20MHz, 40MHz, 60MHz, etc. Candidate subbands define the boundary of frequency selection. Target frequencies refer to specific frequencies within the candidate subbands selected through spectrum scanning and quality assessment; the target frequency is the frequency within the candidate subband that yields the highest throughput.

[0034] Furthermore, the periodic spectrum scanning strategy is determined by judging the channel quality parameters of the current communication link, and is determined through the following steps: Construct the correlation between channel quality parameters and scanning period for each communication link to form a spectrum scanning period library; Based on the channel quality parameters of the current communication link, the corresponding scanning period is matched from the spectrum scanning period library to determine the periodic spectrum scanning strategy.

[0035] By installing a spectrum sensing module on the airborne end, the spectrum scanning interface provided by the industry-grade drone SDK is periodically invoked in the drone to obtain the interference levels of each candidate sub-band within a specified frequency band. Simultaneously, to balance the measurement accuracy and system overhead of the drone's airborne end, a dynamic periodic spectrum scanning strategy is adopted. This strategy determines the scanning period of the spectrum sensing module based on changes in channel quality parameters, such as... Figure 2 As shown in Table 1, channel quality parameters include signal-to-noise ratio (SNR). In a specific example, the relationship between SNR and scan period is shown in Table 1.

[0036] Table 1 Dynamic Periodic Spectrum Scanning Strategy

[0037] The interference level acquired in real time by the spectrum sensing module is the measured interference level value. Due to the influence of the measurement environment, the measured interference level value will have some interference fluctuations. To eliminate interference fluctuations, the measured interference level value can be filtered, as follows:

[0038] in, Let be the interference level filter value of candidate sub-band j in candidate frequency band i. This represents the current interference level measurement value of candidate subband j in candidate frequency band i. This represents the interference level filter value for candidate subband j in candidate frequency band i during the previous scan cycle, where α is the filter coefficient used to balance real-time performance and stability. For example, α∈[0.3,0.7].

[0039] The aforementioned spectrum sensing module acquires the interference level of each candidate sub-band within a specified frequency band range. This can be either a measured interference level or a filtered interference level, and the choice can be made according to actual needs. No limitation is made here.

[0040] By employing a periodic spectrum scanning strategy to achieve adaptive adjustment of the scanning period, the system can automatically adjust the strategy according to environmental changes and always maintain optimal performance.

[0041] In this invention, an industry-grade UAV onboard spectrum sensing interface is used to directly measure the interference level of the actual receiving environment of the UAV. Compared with ground control terminal measurement or static electromagnetic map, the measurement accuracy is improved, providing a more accurate decision basis for frequency point selection.

[0042] Interference level measurements and / or interference level filtered values ​​are associated with the current GPS location and stored to form spectrum measurement data, which is used to construct an electromagnetic environment map.

[0043] The electromagnetic environment map is determined through the following steps: Read spectrum measurement data from multiple drones. The spectrum measurement data includes location data, frequency band information, sub-band information, interference level, signal-to-noise ratio, and timestamp. An interpolation algorithm is used to fuse spectrum measurement data from multiple UAVs to construct an electromagnetic environment map.

[0044] Furthermore, an interpolation algorithm is used to fuse the spectrum measurement data of multiple UAVs to construct an electromagnetic environment map, specifically including: Based on the spectrum measurement data of multiple drones, the location data of the drones at each sampling point is obtained; By analyzing the location data of each sampling point and the distance between the location of the point to be interpolated in the electromagnetic environment map, and combining the Gaussian kernel smoothing coefficient, the measurement weight corresponding to each sampling point is given. Based on the measurement weights corresponding to each sampling point, the spectrum measurement data are fused to obtain the spectrum data of each interpolation point, thus forming an electromagnetic environment map.

[0045] The spectral data of each interpolation point are specifically represented as follows:

[0046] Among them, I map (p) represents the spectral data of the point to be interpolated corresponding to location data p in the electromagnetic environment map, where K is the total number of sampling points for each UAV. k Here, σ represents the location data corresponding to sampling point k, and I represents the Gaussian kernel smoothing coefficient. k For sampling point k at p k The spectral measurement data at the location is given by exp(), which is an exponential function.

[0047] In a specific example, there are five sampling points: k=1, p1=0, I1=10; k=2, p2=2, I2=15; k=3, p3=4, I3=8; k=4, p4=6, I4=20; k=5, p5=8, I5=12. In this example, the spectral data at the point to be interpolated, p=2 meters, is calculated, and σ=1 is taken. map (p)=(0.1353×10+1×15+0.1353×8+0.0003×20+1.52×10 −8 ×12=1.353+15+1.0824+0.006+0) / (0.1353+1+0.1353+0.0003+1.52×10 −8 )≈17.4414 / 1.2709=13.72.

[0048] The spectral data in the electromagnetic environment map includes location data, frequency band information, sub-band information, interference level, signal-to-noise ratio, and timestamp.

[0049] After constructing the electromagnetic environment map, it is divided into multiple regions based on location data, and a throughput correction coefficient is defined for each region. In a specific example, the electromagnetic environment map includes the main transformer region, the GIS region, the switchyard region, and an open area. The throughput correction coefficient for the main transformer region is 0.85, for the GIS region it is 0.9, for the switchyard region it is 0.95, and for the open area it is 1. By identifying the drone's location using an airborne terminal, the throughput correction coefficients for each geographical location are retrieved from the electromagnetic environment map during throughput prediction.

[0050] The UAV's frequency adaptive switching system also includes a cloud-based system. Data collected by the communication quality monitoring module in the ground control unit and the spectrum sensing module in the airborne unit are uploaded to the cloud for data fusion and global optimization. For example, after calibration is completed on the airborne unit, the mapping parameters, UAV model identification, frequency band information, and calibration time are uploaded to the cloud. Before takeoff, the UAV downloads shared calibration data from the cloud based on the model identification to determine the signal power mapping relationship. The cloud can manage the version of the signal power mapping relationship, ensuring that each UAV uses the latest signal power mapping relationship. Simultaneously, the cloud records the effectiveness of each use of the signal power mapping relationship, continuously optimizing data quality.

[0051] The spectrum measurement data used in constructing the electromagnetic environment map is downloaded from the cloud. Once the electromagnetic environment map is completed, it is also uploaded to the cloud so that subsequent drones can download the map directly from the cloud and obtain accurate interference information upon their first entry into each area. Each drone's flight uploads new data, updating the electromagnetic environment map based on the latest data from the cloud to ensure continuous improvement in map accuracy.

[0052] By enabling data sharing and multi-drone collaborative electromagnetic environment map construction through the cloud, multiple drones can share data and interference information, and obtain accurate predictions upon first entering the area, reducing the cost of repeated calibration.

[0053] Furthermore, based on the pre-configured list of backup frequency points for hot backup, channel quality parameters are assessed, and a handover trigger signal is generated, specifically including: If the channel quality parameters of the current communication link are lower than the service threshold, a first handover signal is generated. Based on the first switching signal, the first backup frequency point is determined from the list of backup frequency points pre-configured for the hot backup frequency point, and the throughput of the first backup frequency point is judged to give a switching trigger signal.

[0054] In one specific implementation, when the handover trigger module generates a handover trigger, the airborne decision module directly reads the first backup frequency from the backup frequency list, confirms that the available throughput of the first backup frequency meets the service threshold, generates a handover trigger signal, calls the SDK hot backup handover interface, and performs a seamless handover. Image transmission is uninterrupted during the handover process, and the handover time is <20ms. After the handover is completed, the original operating frequency is downgraded to a backup frequency, and the backup frequency list continues to be updated.

[0055] Furthermore, the list of pre-configured spare frequency points for hot backup frequency points is determined through the following steps: The backup frequency list maintenance unit is used to maintain the first backup frequency, store the frequency identifier and score, and the list structure includes the current working frequency, the first backup frequency and the score of each frequency. The dynamic update unit is used to periodically evaluate all candidate frequency points and set the candidate frequency point with the highest predicted throughput as the first backup frequency point. The pre-switching unit is used to actively call the SDK interface to pre-configure the first backup frequency point when the backup frequency point score exceeds the current working frequency point, but without triggering an actual switch. The seamless switching execution unit is used to call the SDK hot backup switching interface when a switch is triggered to perform a seamless switch.

[0056] Specifically, all candidate frequencies are periodically evaluated, and the candidate frequency with the highest predicted throughput is set as the first backup frequency. This includes: Obtain all current candidate frequency points and form a candidate frequency point list; Based on a preset update cycle, and combining an electromagnetic environment map and a throughput prediction function, the predicted throughput of each candidate frequency point in the candidate frequency point list is calculated. Based on the predicted throughput of each candidate frequency point, sort the candidate frequency points in the candidate frequency point list and update the order of each candidate frequency point in the candidate frequency point list; Set the candidate frequency with the highest predicted throughput from the candidate frequency list as the first backup frequency; Based on the update cycle, the predicted throughput of each candidate frequency point is updated in real time, and the first backup frequency point is dynamically replaced.

[0057] In one specific implementation, all current candidate frequencies are analyzed to form a candidate frequency list, resulting in a candidate frequency pool. A dynamic update unit, based on a preset update cycle and combining an electromagnetic environment map and a throughput prediction function, calculates the predicted throughput achievable by the UAV at each candidate frequency, reflecting the communication transmission capability of each candidate frequency. The lower the interference and the higher the signal-to-noise ratio of a candidate frequency, the higher its throughput and the better its transmission capability. Next, all candidate frequencies are sorted from highest to lowest predicted throughput; higher predicted throughput indicates better communication quality and suitability as a backup. The candidate frequency with the highest predicted throughput is set as the first backup frequency, serving as the preferred seamless switchover option in case of primary frequency failure or excessive interference. The predicted throughput of each candidate frequency is updated in real-time according to the update cycle, dynamically replacing the first backup frequency to ensure that it remains the optimal choice under the current environment.

[0058] The airborne terminal also includes a hot backup management module, which comprises a spare frequency list maintenance unit, a dynamic update unit, a pre-switching unit, and a seamless handover execution unit. This module is used to build, update, and maintain the spare frequency list, ensuring it includes optimal candidate frequencies. The spare frequency list is updated after each spectrum scan and throughput prediction. The latest predicted achievable throughput for each candidate subband in the throughput prediction module is used as the frequency score for each candidate subband. When the frequency score of a candidate frequency in the spare frequency list exceeds the frequency score of the current operating frequency, the SDK interface is proactively invoked to pre-configure the spare frequency without triggering an actual handover, ensuring the spare frequency is ready during handover and achieving proactive pre-configuration. By updating the spare frequency list after each spectrum scan, the timeliness of the candidate frequency information in the spare frequency list is guaranteed.

[0059] By utilizing the hot backup switching function of the industry-grade drone image transmission system, seamless switching is achieved, ensuring uninterrupted image transmission and smooth remote control during the switching process, significantly improving operational continuity and safety.

[0060] Specifically, the signal power mapping relationship is determined through the following steps: Read the actual received power at different frequencies in each frequency band and the signal strength indication value of the drone at different frequencies from the cloud; By establishing the correspondence between the actual received power and the signal strength indication values, data pairs are formed, and a set of collected data points is constructed. Based on the collected data points, the actual received power and signal strength indication values ​​in different frequency bands are fitted to obtain the mapping relationship parameters for different frequency bands; By combining the drone's model identifier, a mapping relationship between the model identifier and different frequency bands is constructed to form a signal power mapping relationship.

[0061] The airborne terminal also includes a signal calibration module, which establishes a mapping relationship between the signal strength indication value returned by the SDK and the actual received power, and determines the signal power mapping relationship.

[0062] In one specific implementation, the signal calibration module establishes a mapping relationship between the signal strength indication value returned by the SDK and the actual received power. First, it downloads shared calibration data of the same model and frequency band from the cloud. If no shared calibration data is available in the cloud, it performs a local rapid calibration, selecting 1-2 test points from each frequency band in the local calibration data to establish a mapping relationship P. est =f(R sdk In a specific example, the signal power mapping relationship is as follows:

[0063] Among them, P est For the actual received power, R sdk is the signal strength indication value, and a and b are mapping parameters.

[0064] During flight, data pairs consisting of signal strength indication values ​​and actual received power are continuously collected. The mapping parameters are then updated online using a recursive least squares method to achieve online adaptive correction, specifically as follows:

[0065] Where n represents the number of times the drone collects data. This is the parameter vector after the nth update, and the parameter vector includes the parameters of each mapping relationship. Let K be the parameter vector after the (n-1)th update. n Let y be the Kalman gain coefficient for the nth update. n This represents the actual received power of the drone during the nth data acquisition. Let P be the signal strength indication value collected by the UAV for the nth time. n Let P be the covariance matrix updated for the nth time. n-1 Let I be the covariance matrix of the (n-1)th update, I be the identity matrix, and λ be the forgetting factor, where λ∈[0.95,0.99]. The forgetting factor is used to balance the weights of historical data and current data.

[0066] By updating the mapping parameters online, the system can automatically adjust its strategy according to environmental changes, always maintaining optimal performance.

[0067] Furthermore, based on the signal power mapping relationship, and combined with the interference level and throughput correction coefficient of each candidate sub-band, the achievable throughput of each candidate sub-band is predicted, specifically including: By using the signal power mapping relationship and combining the signal strength indication values ​​of each frequency band, the actual received power of each frequency band can be determined. Based on the actual received power of each frequency band and the interference level of each candidate sub-band, the effective signal-to-noise ratio of each candidate sub-band is calculated. Based on the effective signal-to-noise ratio of each candidate subband, match the corresponding throughput prediction function; Substitute the effective signal-to-noise ratio of each candidate subband into the corresponding throughput prediction function to obtain the first predicted throughput value of each candidate subband. Based on the current location of the drone, determine the throughput correction factor for the current area; By using a throughput correction factor, the first predicted value of throughput is corrected, and the achievable throughput of each candidate subband is given.

[0068] The throughput prediction module employs a piecewise linear prediction model to achieve lightweight throughput prediction at the airborne end. Through signal power mapping, it provides the actual received power corresponding to the current signal strength indication value, and then calculates the effective signal-to-noise ratio, specifically expressed as follows:

[0069] in, Let be the effective signal-to-noise ratio of candidate subband j in candidate frequency band i. The actual received power of candidate subband j in candidate frequency band i is calculated based on the signal power mapping relationship. The interference level of candidate subband j in candidate frequency band i is obtained by measuring the spectrum sensing module.

[0070] It is important to understand that in the embodiments provided by this invention, the units for the actual received power and interference level are dBm. dBm is a logarithmic decibel system, which is a relative logarithmic dimension of power. When both the actual received power and interference level are in dBm, linear power division is not required; the values ​​are directly subtracted, and the result is the effective signal-to-noise ratio (SNR). In other embodiments, if the units for the actual received power and interference level are not dBm, then the units need to be standardized before calculating the effective SNR.

[0071] Substituting the calculated effective signal-to-noise ratio into the corresponding throughput prediction function, the throughput prediction function is specifically expressed as:

[0072] Among them, T base SNR is the initial throughput prediction value. eff For the effective signal-to-noise ratio, β1 is the first prediction coefficient, T1 is the first prediction constant, γ1 is the first prediction threshold, β2 is the second prediction coefficient, T2 is the second prediction constant, γ2 is the second prediction threshold, β3 is the third prediction coefficient, T3 is the third prediction constant, γ3 is the third prediction threshold, T4 is the fourth prediction constant, γ4 is the fourth prediction threshold, and u(x) is the unit step function.

[0073] In a specific example, γ1=5, γ2=15, γ3=25, γ4=35, β1=1.0, β2=1.5, β3=1.0, T1=5, T2=15, T3=30, T4=40, then the throughput prediction function is:

[0074] Among them, T base SNR is the first predicted value for throughput. eff The effective signal-to-noise ratio (SNR) is used. A preliminary throughput prediction can be made using the throughput prediction function and the effective SNR. Next, the throughput correction coefficient corresponding to the UAV is read from the electromagnetic environment map, and the first predicted throughput value is corrected to obtain the second predicted throughput value, specifically expressed as:

[0075] Among them, T pred T is the second predicted value for throughput. base λ is the first predicted value for throughput. zone This is the throughput correction factor. The second predicted throughput value corresponding to each candidate subband is taken as the achievable throughput of each candidate subband, with reference to... Figure 3 .

[0076] Furthermore, the frequency switching operation is performed by combining the switching trigger signal and the collaborative decision-making strategy, specifically including: Acquire the target frequency point on the airborne end and the first backup frequency point for the switching trigger signal on the ground control end; If the first backup frequency point and the target frequency point are the same, then execute the switching trigger signal to complete the frequency point switching; If the first backup frequency and the target frequency are different, the communication quality is determined based on the channel quality parameters of the current communication link, and the first backup frequency or the target frequency is selected to perform frequency switching.

[0077] In a specific example, if the communication quality reaches the communication threshold, it means that the communication quality can meet the current basic communication needs. The target frequency point is then sent to the ground control station for confirmation, and a frequency switch is performed based on the frequency point result fed back by the ground control station. If the communication quality does not reach the communication threshold, it means that the communication quality cannot meet the current basic communication needs, and an emergency frequency switch is required. Therefore, the target frequency point of the airborne terminal is selected for frequency switch to ensure the fastest recovery of communication quality. That is, the airborne terminal prioritizes its own decision.

[0078] The airborne decision-making module selects the optimal candidate frequency as the target frequency based on the available throughput of each candidate subband, while simultaneously confirming that the available throughput of the target frequency meets the service threshold. If the target frequency meets the service requirements, its information is reported to the ground control station for record-keeping, supporting manual intervention for coverage. Decision information for each frequency switch is recorded, including the decision time, target frequency, and available throughput, for subsequent analysis and optimization.

[0079] Reference Figure 4 By installing an airborne decision-making module, the processing delay at the ground control end can be eliminated, allowing decisions to be made on-board and unaffected by communication delays at the ground control end. Simultaneously, response speed is improved; the system can execute decisions directly upon receiving the handover trigger signal, without waiting for a response from the ground control end. Furthermore, even in the event of a communication interruption with the ground control end, the airborne end can still complete the handover decision, reducing the computational load on the ground control end.

[0080] Reference Figure 5 This invention provides a frequency adaptive switching method for unmanned aerial vehicles (UAVs), the specific steps of which are as follows: The channel quality parameters of the current communication link are monitored in real time. Combined with the list of backup frequency points pre-configured for hot backup frequency points, the channel quality parameters are judged and a handover trigger signal is generated. The interference levels of each candidate subband are collected in real time using a periodic spectrum scanning strategy. Based on the signal power mapping relationship, and combined with the interference level and throughput correction coefficient of each candidate sub-band, the achievable throughput of each candidate sub-band is predicted. The target frequency is determined by the available throughput of each candidate subband, and the frequency switching operation is performed in conjunction with the switching trigger signal.

[0081] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working process of the described steps can be referred to the corresponding process in the aforementioned system embodiments, and will not be repeated here.

[0082] Although preferred embodiments of the invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including both the preferred embodiments and all changes and modifications falling within the scope of the invention. Clearly, those skilled in the art can make various alterations and modifications to the invention without departing from its spirit and scope. Thus, if these modifications and modifications of the invention fall within the scope of the claims and their equivalents, the invention is also intended to include these modifications and modifications.

Claims

1. A frequency adaptive switching system for unmanned aerial vehicles (UAVs), characterized in that, Including ground control and airborne control; The ground control terminal is used to monitor the channel quality parameters of the current communication link in real time. Combined with the list of backup frequency points pre-configured for hot backup frequency points, it judges the channel quality parameters and generates a handover trigger signal. In addition, construct an electromagnetic environment map and record the throughput correction coefficient for each geographical location; The airborne terminal is used to collect the interference level of each candidate sub-band in real time through a periodic spectrum scanning strategy; Furthermore, based on the signal power mapping relationship, and combined with the interference level and throughput correction coefficient of each candidate sub-band, the achievable throughput of each candidate sub-band is predicted. Furthermore, the target frequency is determined by the available throughput of each candidate subband, and the frequency switching operation is performed by combining the switching trigger signal and the collaborative decision-making strategy.

2. The frequency adaptive switching system for UAVs as described in claim 1, characterized in that, An electromagnetic environment map is determined through the following steps: Read spectrum measurement data from multiple drones. The spectrum measurement data includes location data, frequency band information, sub-band information, interference level, signal-to-noise ratio, and timestamp. An interpolation algorithm is used to fuse spectrum measurement data from multiple UAVs to construct an electromagnetic environment map.

3. The frequency adaptive switching system for UAVs as described in claim 2, characterized in that, An interpolation algorithm is used to fuse spectrum measurement data from multiple UAVs to construct an electromagnetic environment map, specifically including: The position data of each drone is obtained based on the spectrum measurement data of multiple drones; Analyze the location data of each UAV and the distance between the location of the point to be interpolated in the electromagnetic environment map, and combine the Gaussian kernel smoothing coefficient to give the corresponding measurement weight for each UAV. Based on the measurement weights of each UAV, the spectrum measurement data are fused to obtain the spectrum data of each interpolation point, thus forming an electromagnetic environment map.

4. The frequency adaptive switching system for a UAV as described in claim 1, characterized in that, Based on the pre-configured list of backup frequency points for hot backup, the channel quality parameters are assessed, and a handover trigger signal is generated, specifically including: Based on a preset hysteresis time window, the channel quality parameters of the current communication link are determined: If the channel quality parameters of the current communication link are lower than the service threshold, a first handover signal is generated. Based on the first switching signal, the first backup frequency point is determined from the list of backup frequency points pre-configured for the hot backup frequency point, and the throughput of the first backup frequency point is judged to give a switching trigger signal.

5. The frequency adaptive switching system for a UAV as described in claim 4, characterized in that, The list of pre-configured backup frequency points for hot backup is determined through the following steps: The backup frequency list maintenance unit is used to maintain the first backup frequency, store the frequency identifier and score, and the list structure includes the current working frequency, the first backup frequency and the score of each frequency. The dynamic update unit is used to periodically evaluate all candidate frequency points and set the candidate frequency point with the highest predicted throughput as the first backup frequency point. The pre-switching unit is used to actively call the SDK interface to pre-configure the first backup frequency point when the backup frequency point score exceeds the current working frequency point, but without triggering an actual switch. The seamless switching execution unit is used to call the SDK hot backup switching interface when a switch is triggered to perform a seamless switch.

6. The frequency adaptive switching system for a UAV as described in claim 5, characterized in that, Periodically evaluate all candidate frequencies and designate the candidate frequency with the highest predicted throughput as the first backup frequency. Specifically, this includes: Obtain all current candidate frequency points and form a candidate frequency point list; Based on a preset update cycle, and combining an electromagnetic environment map and a throughput prediction function, the predicted throughput of each candidate frequency point in the candidate frequency point list is calculated. Based on the predicted throughput of each candidate frequency point, sort the candidate frequency points in the candidate frequency point list and update the order of each candidate frequency point in the candidate frequency point list; Set the candidate frequency with the highest predicted throughput from the candidate frequency list as the first backup frequency; Based on the update cycle, the predicted throughput of each candidate frequency point is updated in real time, and the first backup frequency point is dynamically replaced.

7. The frequency adaptive switching system for a UAV as described in claim 1, characterized in that, The signal power mapping relationship is determined through the following steps: Read the actual received power at different frequencies in each frequency band and the signal strength indication value of the drone at different frequencies from the cloud; By establishing the correspondence between the actual received power and the signal strength indication values, data pairs are formed, and a set of collected data points is constructed. Based on the collected data points, the actual received power and signal strength indication values ​​in different frequency bands are fitted to obtain the mapping relationship parameters for different frequency bands; By combining the drone's model identifier, a mapping relationship between the model identifier and different frequency bands is constructed to form a signal power mapping relationship.

8. The frequency adaptive switching system for a UAV as described in claim 7, characterized in that, Based on the signal power mapping relationship, and combined with the interference level and throughput correction coefficient of each candidate sub-band, the achievable throughput of each candidate sub-band is predicted, specifically including: By using the signal power mapping relationship and combining the signal strength indication values ​​of each frequency band, the actual received power of each frequency band can be determined. Based on the actual received power of each frequency band and the interference level of each candidate sub-band, the effective signal-to-noise ratio of each candidate sub-band is calculated. Based on the effective signal-to-noise ratio of each candidate subband, match the corresponding throughput prediction function; Substitute the effective signal-to-noise ratio of each candidate subband into the corresponding throughput prediction function to obtain the first predicted throughput value of each candidate subband. Based on the current location of the drone, determine the throughput correction factor for the current area; By using a throughput correction factor, the first predicted value of throughput is corrected, and the achievable throughput of each candidate subband is given.

9. The frequency adaptive switching system for a UAV as described in claim 1, characterized in that, The frequency switching operation is performed by combining the switching trigger signal and the collaborative decision-making strategy, specifically including: Acquire the target frequency point on the airborne end and the first backup frequency point for the switching trigger signal on the ground control end; If the first backup frequency point and the target frequency point are the same, then execute the switching trigger signal to complete the frequency point switching; If the first backup frequency and the target frequency are different, the communication quality is determined based on the channel quality parameters of the current communication link, and the first backup frequency or the target frequency is selected to perform frequency switching.

10. A frequency adaptive switching method for unmanned aerial vehicles (UAVs), characterized in that, The frequency adaptive switching system for a UAV as described in any one of claims 1-9 includes: The channel quality parameters of the current communication link are monitored in real time. Combined with the list of backup frequency points pre-configured for hot backup frequency points, the channel quality parameters are judged and a handover trigger signal is generated. The interference levels of each candidate subband are collected in real time using a periodic spectrum scanning strategy. Based on the signal power mapping relationship, and combined with the interference level and throughput correction coefficient of each candidate sub-band, the achievable throughput of each candidate sub-band is predicted. The target frequency is determined by the available throughput of each candidate subband, and the frequency switching operation is performed in conjunction with the switching trigger signal.

Citation Information

Patent Citations

  • Unmanned aerial vehicle communication frequency band distribution system and method based on semi-supervised learning algorithm

    CN113301654A