Radio frequency and cellular cooperative networked uav global monitoring method and system

Abstract

The application discloses a kind of radio frequency and cellular coordination's network connection unmanned plane global monitoring method and system, first in cellular signal blind area, the radio frequency fingerprint and position information of unmanned plane target are acquired by radio frequency sensing equipment and continuously tracked, generate anonymous track and pre-discovery task;When target enters the edge transition zone of base station coverage area, improve radio frequency tracking frequency, start cellular network monitoring module and mark to be associated state;When terminal access event is captured, candidate anonymous track is selected from the pre-discovery task to be associated in combination with the preset space-time association condition;Identity association is confirmed by fixed order progressive hierarchical check;Finally, candidate anonymous track and the real-name track of access terminal are cross-source fusion, and the global continuous flight track with the identity information of access terminal is output.The application breaks technical prejudice, realizes radio frequency and cellular coordination, eliminates blind area, and ensures that unmanned plane is continuously traceable across regional monitoring.

Description

Technical Field

[0001] This application relates to the fields of low-altitude unmanned aerial vehicle (UAV) management, radio monitoring and mobile communication integration technology. Specifically, it relates to a networked UAV full-domain monitoring method and system that deeply integrates radio frequency passive sensing and cellular communication. It is particularly suitable for complex airspaces such as mountainous areas, suburbs, airport perimeters and near seas where there are base station signal blind spots. It can realize continuous monitoring, identification and trajectory tracing of UAVs from signal blind spots to cellular base station signal coverage areas at all times. Background Technology

[0002] With the rapid development of the low-altitude economy, the application of networked drones equipped with 4G / 5G communication modules is becoming increasingly widespread. As a result, violations such as "black flights," intrusion into no-fly zones, and interference with aviation safety are occurring frequently. The industry's demand for monitoring drones with full coverage, identifiable identities, and continuous traceability of flight paths continues to increase.

[0003] Currently, two independent drone monitoring systems have been established within the industry:

[0004] 1. Cellular Network Monitoring Solution: This solution relies on 4G / 5G base stations and the core network to acquire real-name data such as the drone's serial number (SN), network access events, and cell-level positioning. Its advantage lies in obtaining the device's real-name identity information, enabling identification and flight track recording within the base station signal coverage area. However, this solution is completely unable to collect any data in remote airspace, mountainous areas, suburbs, and near-shore areas where base station signals are not available, creating a monitoring blind spot. More critically, when a drone flies from a base station signal blind spot into a base station signal coverage area, the cellular network will identify it as a completely new network access terminal, resulting in the complete loss of flight data during the base station signal blind spot phase. This causes a break in the flight track of the same drone, making full-process traceability impossible.

[0005] 2. Pure RF Detection Solution: This solution captures RF signals from drone remote control and image transmission via an array of RF sensors. It can detect and continuously track drones in base station signal blind spots, generating location tracks. Its advantage is that it does not rely on drone cooperation and can detect targets in advance. However, this solution can only obtain the target's RF fingerprint and anonymous location information, and cannot identify the drone's equipment identity. It can only determine that "a drone exists in the airspace," but cannot complete equipment traceability, resulting in severely insufficient regulatory effectiveness.

[0006] In existing technologies, radio frequency monitoring systems and cellular network monitoring systems are completely separate and operate independently, with fundamental differences in the following three dimensions:

[0007] First, the operating space domains are different: radio frequency monitoring systems operate in base station blind spots, while cellular network monitoring systems operate in base station coverage areas;

[0008] Second, the data formats are different: radio frequency monitoring systems output anonymous tracks, while cellular network monitoring systems output real-name signaling.

[0009] Third, the regulatory objectives are different: radio frequency monitoring systems serve target discovery, while cellular network monitoring systems serve identity recognition.

[0010] This disconnect leads to the following four core flaws:

[0011] 1. Monitoring blind spots: Base station signal blind spots create monitoring vacuums, making it impossible to detect drone activities in advance;

[0012] 2. Track interruption: When the same UAV flies across blind spots and coverage areas, its track is interrupted;

[0013] 3. Identity Loss: Anonymous radio frequency tracks cannot be linked to real-name cellular identities, making them untraceable;

[0014] 4. Target confusion: When multiple drones enter and leave the coverage area at the same time, target mismatch is very likely to occur.

[0015] More importantly, there has long been a technological bias in this field. Researchers generally believe that radio frequency anonymized data and cellular real-name data belong to different data domains and have no correlation value. Therefore, they only optimize the performance of a single system and have no motivation to link the two independent systems in a time sequence, exchange data, and conduct collaborative verification. They have also failed to develop a complete solution for cross-regional track correlation.

[0016] Therefore, there is an urgent need for a comprehensive continuous monitoring solution that can break through the aforementioned technical barriers, eliminate technical biases, and integrate the advantages of radio frequency passive monitoring and cellular network monitoring. Summary of the Invention

[0017] The purpose of this application is to provide a method and system for full-domain monitoring of networked drones that combines radio frequency and cellular technologies, in order to solve the problems of fragmented radio frequency monitoring and cellular monitoring data, inability to correlate heterogeneous cross-source data, and inability to form continuous real-name flight tracks from non-coverage areas to coverage areas in the prior art. At the same time, it aims to overcome the integration barriers and technical biases of the two monitoring systems in this field.

[0018] The first objective of this application is to provide a method for global monitoring of connected drones using a combination of radio frequency and cellular technologies.

[0019] The aforementioned objective of this application is achieved through the following technical solution:

[0020] A method for comprehensive monitoring of connected drones using a combination of radio frequency and cellular technologies, the method comprising:

[0021] S1, In areas not covered by cellular base station signals, passive sensing is performed on the UAV target through radio frequency sensing equipment to obtain the radio frequency fingerprint and location information of the UAV target and continuously track it to form an anonymous flight path, and a corresponding pre-discovery task is created based on the anonymous flight path;

[0022] S2, monitor the location of the drone target corresponding to the pre-discovery task; when it is determined that the drone target has reached the transition zone at the edge of the base station coverage area, increase the radio frequency tracking frequency, start the cellular network monitoring module corresponding to the base station coverage area, and mark the pre-discovery task as pending association.

[0023] S3, when the cellular network monitoring module captures a terminal access event within the base station coverage area, it obtains the identity information and access spatiotemporal information of the access terminal, and selects candidate anonymous tracks from the pre-discovery tasks marked as pending association based on preset spatiotemporal association conditions.

[0024] S4, perform a fixed-order progressive hierarchical verification between the UAV target corresponding to the candidate anonymous flight path and the access terminal to confirm the identity association, wherein the progressive hierarchical verification includes at least a first-level verification based on radio frequency fingerprint matching and a second-level verification based on kinematic characteristics verification.

[0025] S5, in response to identity association confirmation, the candidate anonymous flight track is fused with the real-name flight track of the access terminal in the cellular base station signal coverage area, and the full-domain continuous flight track with the identity information of the access terminal is output.

[0026] Preferably, S1 specifically includes:

[0027] S11, a radio frequency sensor array is deployed in a distributed manner in the area not covered by the cellular base station signal, the hardware and data link of the radio frequency sensor array being independent of the cellular network;

[0028] S12, capture the UAV radio frequency signal through the radio frequency sensor array, extract the physical layer features of the UAV radio frequency signal including carrier frequency offset, phase noise and protocol type, and construct a unique radio frequency fingerprint of the UAV target;

[0029] S13, The real-time coordinates of the UAV target are calculated using a time difference of arrival (TDOA) positioning algorithm or an angle of arrival (AGA) positioning algorithm, and used as the location information;

[0030] S14, assign a unique radio frequency (RF) identifier to the drone target, and create a corresponding pre-discovery task based on the RF identifier, and store the RF fingerprint, the location information and the discovery time in the pre-discovery task;

[0031] S15, continuously collect the position data of the UAV target at a fixed sampling period, and accumulate them into a continuous anonymous flight track according to the time series, and store it in the pre-discovery task library.

[0032] Preferably, in step S2, determining that the drone target has reached the transition zone at the edge of the base station coverage area specifically includes:

[0033] S21, Pre-calibrate the signal coverage boundary coordinates of each cellular base station and establish a base station coverage database;

[0034] S22, call the base station coverage database to calculate the shortest distance between the real-time coordinates of the UAV target and the signal coverage boundary closest to the real-time coordinates of the UAV target;

[0035] S23, when the shortest distance is less than the preset edge distance threshold, it is determined that the drone target has reached the edge transition zone of the base station coverage area.

[0036] Preferably, in step S3, the step of selecting candidate anonymous tracks from the pre-discovery tasks marked as pending association, based on preset spatiotemporal correlation conditions, specifically includes:

[0037] S31, based on the access time of the access terminal, set a time window and filter out the pre-discovery tasks whose start time of the anonymous track falls within the time window;

[0038] S32, extract the last location point of the UAV target in the filtered pre-discovery task within the transition zone at the edge of the base station coverage area, and calculate the spatial distance between the last location point and the geographical coordinates of the cell where the access terminal is located;

[0039] S33, when the spatial distance is less than a preset spatial distance threshold, the anonymous track corresponding to the pre-discovery task is listed as a candidate anonymous track.

[0040] Preferably, in step S4, the progressive hierarchical verification further includes a third-level verification based on device identity information confirmation. Specifically, performing a fixed-order progressive hierarchical verification on the UAV target corresponding to the candidate anonymous flight path and the access terminal to confirm identity association includes:

[0041] S41, the first-level verification based on radio frequency fingerprint matching, includes:

[0042] S411, retrieve the radio frequency fingerprint corresponding to the candidate anonymous track;

[0043] S412, Obtain the radio frequency capability field information reported by the access terminal to the cellular network;

[0044] S413, The feature similarity algorithm is used to compare the radio frequency fingerprint with the radio frequency capability field information in multiple dimensions to obtain a comprehensive matching degree value. The comparison dimensions include at least the carrier frequency offset value and the phase noise spectrum feature.

[0045] S414, when the comprehensive matching degree value is greater than the preset feature matching threshold, it is determined that the first level of verification has passed and the second level of verification has been entered; otherwise, it is determined that the identity association verification has failed and the current identity association verification process is terminated.

[0046] S42, the second-level verification based on kinematic properties, includes:

[0047] S421, calculate the spatial distance between the end point of the candidate anonymous track and the geographical coordinates of the cell where the access terminal is located. When the spatial distance is less than the preset spatial distance threshold, the spatial consistency verification is passed; otherwise, the identity association verification process is terminated.

[0048] S422, calculate the speed difference between the average speed of the UAV target corresponding to the candidate anonymous flight path during the radio frequency tracking phase and the instantaneous speed of the access terminal at the moment of network access. When the speed difference is less than the preset speed difference threshold, it is determined that the speed continuity verification is passed; otherwise, the identity association verification process is terminated.

[0049] S423, perform splicing smoothness verification on the candidate anonymous track and the real-name track of the access terminal in the cellular base station signal coverage area. When the track curvature change rate at the splicing point is less than the preset curvature change rate threshold, the track smoothness verification is passed; otherwise, the identity association verification process is terminated.

[0050] S424, when all the spatial consistency check, speed continuity check and splicing smoothness check are passed, the second level of check is passed and the third level of check is entered;

[0051] S43, the third-level verification based on device identity information, includes:

[0052] S431, retrieve the historical archive of serial numbers of all unmanned aerial vehicles (UAVs) in the entire domain;

[0053] S432, retrieve whether the device serial number of the access terminal exists in the historical archive;

[0054] S433, if the device serial number of the access terminal exists in the historical archive, then the access terminal is marked as a known device, and the drone target corresponding to the candidate anonymous flight path is confirmed to be associated with the access terminal's identity; if it does not exist, then a new file is automatically created for the device serial number, and the drone target corresponding to the candidate anonymous flight path is confirmed to be associated with the access terminal's identity.

[0055] Preferably, in step S423, the step of performing a smoothness check on the splicing of the candidate anonymous flight track and the real-name flight track of the access terminal within the cellular base station signal coverage area specifically includes:

[0056] A spline interpolation algorithm is used to perform curve fitting between the N position points at the end of the candidate anonymous track and the M position points at the beginning of the real track, where N and M are both positive integers greater than or equal to 3;

[0057] Calculate the rate of change of curvature of the fitted curve at the splice point;

[0058] When the rate of curvature change is less than a preset rate of curvature change threshold, the trajectory smoothness condition is determined to be met.

[0059] Preferably, in step S5, the cross-source fusion of the candidate anonymous flight track and the real-name flight track of the access terminal within the cellular base station signal coverage area specifically includes:

[0060] S51, identify the differences between the candidate anonymous track and the real track in terms of coordinate reference and timestamp granularity;

[0061] S52, using the moment of identity association confirmation as the splicing point, the coordinates of the candidate anonymous track and the coordinates of the real track are uniformly transformed to the same coordinate system to achieve spatial reference alignment;

[0062] S53, unify the timestamps of the candidate anonymous tracks and the timestamps of the real-name tracks to the same time base to achieve timeline alignment;

[0063] S54 uses the Kalman filter algorithm to smooth the transition of the track at the splicing point;

[0064] S55, the spliced ​​complete flight track is bound and stored with the device serial number of the access terminal.

[0065] Preferably, the radio frequency and cellular coordinated networked drone all-domain monitoring method further includes:

[0066] S6 When multiple drone targets arrive at the edge transition zone of the base station coverage area at the same time and there are multiple terminal access events, the preset multi-target anti-confusion processing strategy is executed.

[0067] Preferably, in step S6, the preset multi-target anti-obfuscation processing strategy includes the following steps:

[0068] S61, construct the correlation scoring matrix, where the matrix rows correspond to each candidate anonymous track and the matrix columns correspond to each access terminal;

[0069] S62, calculate the comprehensive matching score for each candidate anonymous track-access terminal pair based on radio frequency fingerprint similarity and kinematic matching degree;

[0070] S63, select the candidate anonymous track-access terminal pair with the highest global score and the comprehensive matching score not lower than the preset matching threshold to complete the one-to-one identity binding, and remove the bound candidate anonymous track and access terminal from the matrix;

[0071] S64, repeat steps S62 to S63 until all candidate anonymous tracks and access terminals in the matrix have been bound or there are no matching objects;

[0072] S65 marks candidate anonymous track-access terminal pairs whose comprehensive matching scores are lower than the preset matching threshold as pending manual review.

[0073] The second objective of this application is to provide a networked unmanned aerial vehicle (UAV) all-domain monitoring system that combines radio frequency and cellular technologies.

[0074] The second objective of this application is achieved through the following technical solution:

[0075] A network-connected unmanned aerial vehicle (UAV) all-domain monitoring system that combines radio frequency and cellular technologies, the system comprising:

[0076] The radio frequency sensing module is deployed in areas not covered by cellular base station signals. It is configured to perform passive sensing of UAV targets, acquire the radio frequency fingerprint and location information of the UAV targets and continuously track them to form anonymous tracks, and create corresponding pre-discovery tasks based on the anonymous tracks.

[0077] The transition detection module is configured to monitor the location of the UAV target corresponding to the pre-discovery task. When it is determined that the UAV target has reached the transition zone at the edge of the base station coverage area, the radio frequency tracking frequency is increased, the cellular network monitoring module corresponding to the base station coverage area is activated, and the pre-discovery task is marked as pending association.

[0078] The cellular network monitoring module is configured to capture terminal access events within the coverage area of ​​the base station and obtain the identity information and access spatiotemporal information of the accessing terminal.

[0079] The track filtering module is configured to filter candidate anonymous tracks from pre-discovery tasks marked as pending association based on preset spatiotemporal association conditions.

[0080] The identity association verification module is configured to perform a fixed-order progressive hierarchical verification between the UAV target corresponding to the candidate anonymous flight path and the access terminal to confirm the identity association. The progressive hierarchical verification includes at least a first-level verification based on radio frequency fingerprint matching and a second-level verification based on kinematic characteristics verification.

[0081] The cross-source fusion output module is configured to, in response to identity association confirmation, perform cross-source fusion of the candidate anonymous flight track and the real-name flight track of the access terminal within the cellular base station signal coverage area, and output a full-domain continuous flight track with the identity information of the access terminal.

[0082] Preferably, the radio frequency sensing module includes:

[0083] A distributed radio frequency sensor array deployed in areas not covered by the cellular base station signal, wherein the hardware and data link of the radio frequency sensor array are independent of the cellular network;

[0084] The radio frequency sensor array is configured to: capture UAV radio frequency signals; extract physical layer features from the UAV radio frequency signals, including carrier frequency offset, phase noise, and protocol type, to construct a unique radio frequency fingerprint of the UAV target; calculate the real-time coordinates of the UAV target using a time difference of arrival (TDOA) positioning algorithm or an angle of arrival (AOA) positioning algorithm, as the location information; assign a unique radio frequency identifier to the UAV target, and create a corresponding pre-discovery task based on the unique radio frequency identifier; associate and store the radio frequency fingerprint, the location information, and the discovery time in the pre-discovery task; continuously collect the location data of the UAV target at a fixed sampling period, and accumulate it into a continuous anonymous flight track according to the time series, storing it in the pre-discovery task library.

[0085] Preferably, the transition detection module is specifically configured as follows:

[0086] The signal coverage boundary coordinates of each cellular base station are pre-calibrated to establish a base station coverage area database; the base station coverage area database is called to calculate the shortest distance between the real-time coordinates of the UAV target and the signal coverage boundary closest to the real-time coordinates of the UAV target; when the shortest distance is less than a preset edge distance threshold, it is determined that the UAV target has reached the edge transition zone of the base station coverage area, and the radio frequency tracking frequency is increased.

[0087] Preferably, the track filtering module is specifically configured as follows:

[0088] Based on the access time of the access terminal, a time window is set to filter out pre-discovery tasks whose start time of anonymous tracks falls within the time window; the last location point of the UAV target in the selected pre-discovery task within the transition zone at the edge of the base station coverage area is extracted, and the spatial distance between the last location point and the geographical coordinates of the cell where the access terminal is located is calculated; when the spatial distance is less than a preset spatial distance threshold, the anonymous track corresponding to the pre-discovery task is listed as a candidate anonymous track.

[0089] Preferably, the identity association verification module is specifically configured as follows:

[0090] Performing the first-level verification based on radio frequency fingerprint matching includes: retrieving the radio frequency fingerprint corresponding to the candidate anonymous track; obtaining the radio frequency capability field information reported by the access terminal to the cellular network; using a feature similarity algorithm to compare the radio frequency fingerprint with the radio frequency capability field information in multiple dimensions to obtain a comprehensive matching degree value, wherein the comparison dimensions include at least the carrier frequency offset value and the phase noise spectrum feature; when the comprehensive matching degree value is greater than a preset feature matching threshold, it is determined that the first-level verification has passed and the second-level verification has begun; otherwise, it is determined that the identity association verification has failed and the current identity association verification process has been terminated.

[0091] The second-level verification based on kinematic characteristics includes: calculating the spatial distance between the end point of the candidate anonymous track and the geographic coordinates of the cell where the access terminal is located; if the spatial distance is less than a preset spatial distance threshold, the spatial consistency verification is passed; calculating the speed difference between the average speed of the UAV target corresponding to the candidate anonymous track during the radio frequency tracking phase and the instantaneous speed of the access terminal at the moment of network access; if the speed difference is less than a preset speed difference threshold, the speed continuity verification is passed; performing a stitching smoothness verification on the candidate anonymous track and the real-name track of the access terminal within the cellular base station signal coverage area; if the rate of change of track curvature at the stitching point is less than a preset rate of change of curvature threshold, the track smoothness verification is passed; if the spatial consistency verification, speed continuity verification, and stitching smoothness verification are all passed, the second-level verification is passed, and the process proceeds to the third-level verification.

[0092] Perform a third-level verification based on device identity information, including: retrieving the historical archive of UAV device serial numbers across the entire domain; checking whether the device serial number of the access terminal exists in the historical archive; if the device serial number of the access terminal exists in the historical archive, then mark the access terminal as a known device and confirm that the UAV target corresponding to the candidate anonymous flight path is associated with the access terminal's identity; if it does not exist, then automatically create a new file for the device serial number and confirm that the UAV target corresponding to the candidate anonymous flight path is associated with the access terminal's identity.

[0093] Preferably, the cross-source fusion output module is specifically configured as follows:

[0094] The system identifies the differences between the candidate anonymous flight track and the real-name flight track in terms of coordinate reference and timestamp granularity; using the moment of identity confirmation as the splicing point, it uniformly transforms the coordinates of the candidate anonymous flight track and the real-name flight track to the same coordinate system to achieve spatial reference alignment; it unifies the timestamps of the candidate anonymous flight track and the real-name flight track to the same time reference to achieve time axis alignment; it uses a Kalman filter algorithm to perform smooth transition processing on the flight track at the splicing point; it binds and stores the spliced ​​complete flight track with the device serial number of the access terminal, and outputs a full-domain continuous flight track with the identity information of the access terminal.

[0095] Preferably, the identity association verification module is further configured as follows:

[0096] When multiple drone targets simultaneously arrive at the edge of the base station coverage area and correspond to multiple terminal access events, a preset multi-target anti-aliasing processing strategy is executed, including:

[0097] Construct an association scoring matrix, where rows correspond to candidate anonymous tracks and columns correspond to access terminals. Calculate the comprehensive matching score for each candidate anonymous track-access terminal pair based on RF fingerprint similarity and kinematic matching degree. Select the candidate anonymous track-access terminal pair with the highest global score and a comprehensive matching score not lower than a preset matching threshold to complete one-to-one identity binding, and remove the bound candidate anonymous tracks and access terminals from the matrix. Repeat the above selection and binding operations until all candidate anonymous tracks and access terminals in the matrix are bound or there are no matching objects. Mark candidate anonymous track-access terminal pairs with comprehensive matching scores lower than the preset matching threshold as awaiting manual review.

[0098] Compared with the prior art, this application has the following beneficial effects:

[0099] 1. By relying on independent radio frequency sensing, drone activities can be detected in advance in areas not covered by cellular base station signals, breaking through the limitation that cellular networks can only monitor in the coverage area and achieving full airspace monitoring without blind spots.

[0100] 2. It features a unique three-layer progressive verification architecture with a fixed order, which uses hardware fingerprints, motion characteristics, and official records to match step by step, achieving seamless association between anonymous tracks and real-name identities, ensuring complete and continuous cross-regional tracks with full traceability.

[0101] 3. In response to the current situation where the hardware and protocols of the radio frequency monitoring system and the cellular network monitoring system are isolated, we design isolated storage and standardized data interaction rules to break down technical barriers and industry technical biases and achieve deep collaboration between heterogeneous systems.

[0102] 4. A weighted matching algorithm is adopted to customize a multi-target anti-confusion processing strategy for drone swarm scenarios, effectively avoiding multi-target mismatch and omission.

[0103] 5. Using the device's inherent RF fingerprint as the first-level verification basis, the hardware characteristics cannot be tampered with. Combined with a multi-layer verification mechanism, the overall anti-interference capability is far superior to a single comparison scheme. Attached Figure Description

[0104] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0105] Figure 1 This is a flowchart illustrating a method for full-domain monitoring of a networked drone using a combination of radio frequency and cellular technologies, as described in one embodiment of this application.

[0106] Figure 2 This is a schematic diagram of the structure of a networked drone all-domain monitoring system that combines radio frequency and cellular technologies in one embodiment of this application. Detailed Implementation

[0107] To enable those skilled in the art to better understand the technical solutions in this application, the technical solutions in the embodiments of this application will be clearly and completely described below. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0108] In the embodiments provided in this application, it should be understood that the disclosed methods and systems can be implemented in other ways. The system embodiments described below are merely illustrative. For example, the division of units and modules is only a logical functional division, and in actual implementation, there may be other division methods, such as: multiple units or modules can be combined, or integrated into another system, or some features can be ignored or not executed. In addition, the coupling, direct coupling, or communication connection between the various components shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or modules, and can be electrical, mechanical, or other forms.

[0109] In addition, each functional unit in the various embodiments of this application can be integrated into a single processor, or each unit can be a separate device, or two or more units can be integrated into a single device; each functional unit in the various embodiments of this application can be implemented in hardware or in the form of hardware plus software functional units.

[0110] Those skilled in the art will understand that all or part of the steps of the following method embodiments can be implemented by program instructions and related hardware. The aforementioned program instructions can be stored in a computer-readable storage medium. When the program instructions are executed, they perform the steps of the following method embodiments. The aforementioned storage medium includes various media capable of storing program code, such as mobile storage devices, read-only memory (ROM), magnetic disks, or optical disks.

[0111] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this application, "a plurality of" or "several" means two or more, unless otherwise explicitly specified.

[0112] To enable those skilled in the art to better understand the technical solutions in this application, the technical solutions in the embodiments of this application will be clearly and completely described below. It should be noted that the core innovative ideas of this application can be implemented through a variety of specific technical means.

[0113] like Figure 1 As shown in the figure, this application provides a method for full-domain monitoring of networked drones using a combination of radio frequency and cellular technologies. This method may include the following steps:

[0114] S1, in areas not covered by cellular base station signals, performs passive sensing of the UAV target through radio frequency sensing equipment, obtains the radio frequency fingerprint and location information of the UAV target and continuously tracks it, forming an anonymous flight path, and creates a corresponding pre-discovery task based on the anonymous flight path.

[0115] Specifically, in this embodiment, the radio frequency (RF) sensing device is deployed in areas where cellular base station signals cannot cover (base station blind spots), such as mountainous areas, suburbs, and coastal areas. The RF sensing device employs a passive sensing method, passively receiving RF signals radiated by aerial targets without actively emitting any detection signals. When a UAV flies in a base station blind spot, its remote control and image transmission links radiate RF signals. The RF sensing device captures these signals and extracts the physical layer features carried within them. These physical layer features are determined by the nonlinear characteristics of the transmitter hardware, possessing uniqueness and long-term stability, thus constituting the RF fingerprint of the UAV device. Simultaneously, the RF sensing device uses multiple sensor nodes to measure the time difference or angle of arrival of the same signal, and calculates the UAV's three-dimensional spatial coordinates as its position information using a geometric positioning algorithm. The RF sensing device continuously tracks the UAV, collecting new position data at fixed intervals, accumulating it over time to form a complete anonymous flight track. This anonymous flight track records the UAV's entire flight path in the blind spot, but the device's identity is still unknown. Based on this anonymous flight track, a corresponding pre-discovery task is created for subsequent identity association processes.

[0116] S2, monitor the location of the drone target corresponding to the pre-discovery task, and when it is determined that the drone target has reached the transition zone at the edge of the base station coverage area, increase the radio frequency tracking frequency, start the cellular network monitoring module corresponding to the base station coverage area, and mark the pre-discovery task as pending association.

[0117] Specifically, in this embodiment, the location information of the UAV target corresponding to the pre-discovery task created in S1 is continuously monitored. The system pre-stores the geographical coordinates of the signal coverage boundaries of each cellular base station. As the UAV flies towards the base station coverage area, the distance between the UAV's current position and the nearest base station coverage boundary is calculated in real time. When this distance is less than a preset edge distance threshold, it is determined that the UAV target has reached the edge transition zone of the base station coverage area, that is, the critical area where the UAV is about to enter the cellular network coverage area from the base station blind spot. In response to this determination, three operations are performed: First, the radio frequency tracking frequency is increased, that is, the position sampling density is increased on the basis of the normal tracking frequency to obtain higher accuracy track data for subsequent association; Second, the cellular network monitoring module corresponding to the base station coverage area is activated to start listening to the cellular network signaling in the area; Third, the pre-discovery task is marked as pending association, indicating that the task has entered the identity association preparation stage.

[0118] S3, when the cellular network monitoring module captures a terminal access event within the base station coverage area, it obtains the identity information and access spatiotemporal information of the access terminal, and selects candidate anonymous tracks from the pre-discovery tasks marked as pending association based on preset spatiotemporal association conditions.

[0119] Specifically, in this embodiment, the cellular network monitoring module continuously monitors terminal access events within the base station coverage area. When a drone flies into the base station coverage area, its built-in communication module initiates a network attachment process, at which point a terminal access event is generated on the cellular network side. After capturing this terminal access event, the cellular network monitoring module extracts two types of information: first, the identity information of the access terminal, i.e., the device serial number, used to uniquely identify the device; and second, the access spatiotemporal information of the access terminal, including the access time and cell-level positioning location. After obtaining the above information, it filters from all pre-discovery tasks marked as pending association, based on preset spatiotemporal association conditions. The spatiotemporal association conditions include time window constraints and spatial window constraints: the time window constraint uses the access time as a benchmark, filtering pre-discovery tasks whose difference between the discovery time and the access time is within a preset range; the spatial window constraint uses the cell-level positioning location as a benchmark, filtering pre-discovery tasks whose spatial distance between the last radio frequency tracking location and the cell-level positioning location is within a preset range. Anonymous tracks corresponding to pre-discovery tasks that simultaneously satisfy both time window and spatial window constraints are listed as candidate anonymous tracks and enter the subsequent verification stage.

[0120] S4. Perform a fixed-order progressive hierarchical verification on the drone target corresponding to the candidate anonymous flight path and the access terminal to confirm the identity association. The progressive hierarchical verification includes at least a first-level verification based on radio frequency fingerprint matching and a second-level verification based on kinematic characteristics verification.

[0121] Specifically, in this embodiment, the candidate anonymous flight tracks selected by S3 are used to verify the identity association between the candidate and the access terminal. This verification adopts a fixed-order, progressive, hierarchical verification architecture, which includes at least two levels. The first level of verification is based on radio frequency fingerprint matching: the radio frequency fingerprint corresponding to the candidate anonymous flight track is retrieved, and the radio frequency capability field information reported by the access terminal to the cellular network is obtained. A feature similarity algorithm is used to compare the two in multiple dimensions, including at least the comparison of carrier frequency offset value and phase noise spectrum features, to determine whether the UAV target corresponding to the candidate anonymous flight track and the access terminal are the same terminal device. The second level of verification is based on kinematic characteristic verification: the spatial distance between the end position point of the candidate anonymous flight track and the geographical coordinates of the cell where the access terminal is located is further compared; the speed difference between the average speed of the UAV target corresponding to the candidate anonymous flight track during the radio frequency tracking phase and the instantaneous speed of the access terminal at the moment of network access is compared; and the splicing smoothness comparison between the candidate anonymous flight track and the real-name flight track of the access terminal within the cellular base station signal coverage area is performed, to further verify whether the UAV target corresponding to the candidate anonymous flight track and the access terminal are the same terminal device. The verification at each level proceeds in a fixed order. Only after passing the verification at the previous level can the process proceed to the next level; if any verification fails, the association process is immediately terminated. After all verifications pass, it is confirmed that the drone target corresponding to the candidate anonymous flight path is associated with the identity of the access terminal (i.e., the two are the same terminal device).

[0122] S5, in response to identity association confirmation, performs cross-source fusion of candidate anonymous flight tracks and real-name flight tracks of access terminals within the cellular base station signal coverage area, and outputs a full-domain continuous flight track with access terminal identity information.

[0123] Specifically, in this embodiment, in response to the completion of identity association confirmation in S4, cross-source track fusion is performed. Differences in coordinate reference and timestamp granularity between the candidate anonymous track output from the radio frequency sensing side and the real-name track output from the cellular network side are identified. The radio frequency side track may use a local coordinate system and local clock, while the cellular side track uses a global coordinate system and network time. Both are uniformly transformed to the same coordinate system and time reference to achieve spatial and temporal alignment. At the splicing point, a filtering algorithm is used to smooth the track transition, eliminating track jumps or discontinuities that may occur due to data source switching. After fusion, the spliced ​​complete track is bound and stored with the device serial number of the access terminal, and a continuous flight track with the device's real-name identifier is output according to a standard data format. This track covers the complete flight path from the blind spot to the coverage area, with identity information throughout the entire journey.

[0124] This embodiment of the radio frequency and cellular collaborative network-connected drone full-domain monitoring method achieves deep collaboration between the radio frequency sensing system and the cellular network monitoring system. The entire method follows the logical main line of "blind zone perception, edge preparation, network entry triggering, layered verification, and cross-source fusion". In the blind zone stage, the radio frequency sensing device independently completes the detection, positioning, and continuous tracking of the drone, extracts the device's radio frequency fingerprint, and accumulates anonymous flight tracks, filling the coverage gaps of the cellular network and effectively extending the monitoring range from the coverage area to the blind zone, fundamentally eliminating the monitoring blind zone. In the transition stage, the system senses that the drone is about to enter the coverage area, improves the tracking accuracy in advance, and starts cellular listening to prepare data and processes for subsequent identity association. In the network entry triggering stage, the cellular network monitoring module captures the terminal's access event and uses spatiotemporal constraints to quickly filter candidate targets from multiple pre-discovery tasks, significantly narrowing the scope that needs further verification. In the identity association confirmation stage, a fixed order is used to confirm the identity. The progressive layered verification first compares the hardware-level radio frequency fingerprint, then verifies the kinematic characteristics, and confirms at each level whether the candidate anonymous flight track and the access terminal belong to the same device. It uses the radio frequency fingerprint, which is difficult to forge due to the nonlinear characteristics of the transmitter hardware, as the first line of defense. Then, it is supplemented by spatiotemporal consistency constraints for step-by-step filtering, which greatly reduces the probability of erroneous association at each level. The overall mismatch rate is much lower than that of the single-dimensional comparison scheme, which significantly improves the matching accuracy and anti-interference capability. In the fusion output stage, the cross-source heterogeneous flight track data is spatiotemporally aligned and smoothly stitched together, so that the historical flight information lost in the blind spot can be recovered and bound to the real name identity. Finally, the complete real-name flight track with continuous identity, full coverage, and traceability is output.

[0125] In one embodiment, S1 specifically includes:

[0126] S11 involves the distributed deployment of radio frequency sensor arrays in areas not covered by cellular base station signals. The hardware and data links of the radio frequency sensor arrays are independent of the cellular network.

[0127] Specifically, in this embodiment, multiple high points are selected within the target monitoring area, such as building rooftops, ridges, and towers, to distribute radio frequency sensor nodes. These nodes are connected via wired or wireless means to form a sensor array. Each sensor node includes a wideband antenna, a radio frequency front-end, and a signal processing unit, capable of covering the communication frequency bands commonly used by UAVs. The hardware, power supply system, and data backhaul links of the entire sensor array operate independently of the cellular network infrastructure, without relying on base station clock synchronization or data transmission services, ensuring normal operation even in areas where cellular networks are completely absent.

[0128] S12 captures the UAV's radio frequency signals through a radio frequency sensor array, extracts physical layer features from the UAV's radio frequency signals including carrier frequency offset, phase noise, and protocol type, and constructs a unique radio frequency fingerprint of the UAV target.

[0129] Specifically, in this embodiment, the radio frequency sensor array continuously scans the spatial spectrum. Upon capturing a suspected UAV radio frequency signal, it performs in-depth analysis. Three physical layer features are extracted from the signal: carrier frequency offset, phase noise, and protocol type. The carrier frequency offset is caused by the frequency deviation of the transmitter's crystal oscillator; phase noise is generated by the random phase jitter of the transmitter's local oscillator; and the protocol type is determined by the frame structure and modulation method of the UAV's communication protocol stack. These features collectively constitute the radio frequency fingerprint of the UAV device. Due to manufacturing tolerances and individual differences in the hardware components of different UAV devices, the radio frequency fingerprint is unique to each device and can be used to distinguish between different devices.

[0130] S13 uses a time difference of arrival (TDOA) or angle of arrival (AGA) positioning algorithm to calculate the real-time coordinates of the UAV target as its location information.

[0131] Specifically, in this embodiment, a distributed radio frequency sensor array is used for multi-node joint localization of the same UAV signal. When using the time difference of arrival (TDOA) positioning algorithm, each sensor node records the timestamp of the same signal arriving at its own node. By calculating the time difference of arrival at different nodes, a hyperbolic equation system is constructed to solve for the UAV's position coordinates. When using the angle of arrival (AHA) positioning algorithm, each node measures the direction angle of arrival of the signal through its antenna array. The intersection of multiple direction angles is used to determine the UAV's position. The two algorithms can be selected according to the actual deployment scenario, or they can be combined to improve positioning accuracy.

[0132] S14 assigns a unique radio frequency (RF) identifier to the drone target and creates a corresponding pre-discovery task based on the RF identifier, storing the RF fingerprint, location information, and discovery time in the pre-discovery task.

[0133] Specifically, in this embodiment, a globally unique identifier on the radio frequency (RF) side is assigned to the currently detected drone target, serving as its internal code name during subsequent tracking. Based on this unique RF identifier, a corresponding pre-discovery task is created. The RF fingerprint extracted in S12, the location information calculated in S13, and the discovery time (the moment when the drone target is first discovered within the entire target monitoring area, i.e., the moment when the RF sensing device first captures the drone RF signal corresponding to the drone target in an area not covered by cellular base station signals) are associated and stored in the pre-discovery task, making the pre-discovery task a complete information carrier of the drone target in the blind zone.

[0134] S15 continuously collects the location data of UAV targets at a fixed sampling period, and accumulates them into continuous anonymous tracks in a time series, which are then stored in the pre-discovery task database.

[0135] Specifically, in this embodiment, the radio frequency sensor array continuously locates the UAV target according to a set fixed sampling period, constantly generating new position data points. These position data points are arranged in chronological order to form a continuous trajectory line, i.e., an anonymous flight track. This anonymous flight track records the complete flight path of the UAV in the base station blind zone. The continuously updated anonymous flight track is stored in the pre-discovery task library and associated with the corresponding pre-discovery task.

[0136] In this embodiment, the hardware and data link of the radio frequency sensor array are independent of the cellular network, enabling the system to independently complete detection, positioning, and tracking in areas completely without cellular network coverage, thus ensuring the base station's blind zone monitoring capability from a physical perspective. The radio frequency fingerprint, based on inherent hardware characteristics such as carrier frequency offset and phase noise, is unaffected by environmental changes, possessing long-term stability and device uniqueness, providing a reliable foundation for subsequent identity association. The pre-discovery task associates and stores the radio frequency fingerprint, location information, and discovery time, making each pre-discovery task a complete information archive of the UAV target, providing a structured data foundation for subsequent spatiotemporal screening and hierarchical verification. Anonymous tracks are continuously accumulated at a fixed sampling period, enriching the track data in the pre-discovery task as the tracking process progresses, providing sufficient data points for subsequent kinematic characteristic verification.

[0137] In one embodiment, step S2, determining that the drone target has reached the transition zone at the edge of the base station coverage area, specifically includes:

[0138] S21, Pre-calibrate the signal coverage boundary coordinates of each cellular base station and establish a base station coverage database.

[0139] Specifically, in this embodiment, before or during the initial deployment phase within the monitoring area, all cellular base stations in the area are surveyed. The survey includes parameters such as the latitude and longitude coordinates, antenna height, antenna azimuth, downtilt angle, and transmit power of each base station. Based on these parameters, a standard propagation model is used to calculate the signal coverage area of ​​each base station, generating a set of geographic coordinates for the coverage boundaries. The coverage boundary coordinates of all base stations are then compiled to form a base station coverage area database, which is stored in the system for subsequent real-time querying.

[0140] S22, call the base station coverage database to calculate the shortest distance between the real-time coordinates of the UAV target and the nearest signal coverage boundary to the real-time coordinates of the UAV target.

[0141] Specifically, in this embodiment, during the drone target tracking process, its current coordinates are acquired in real time, and the base station coverage database established in S21 is accessed. The distance between the drone target coordinates and the coverage boundary of each base station in the database is calculated, and the minimum distance value is found, which is the shortest distance from the drone to the nearest base station coverage boundary. This distance value reflects the drone's current distance from the cellular network coverage area.

[0142] S23, when the shortest distance is less than the preset edge distance threshold, it is determined that the drone target has reached the edge transition zone of the base station coverage area.

[0143] Specifically, in this embodiment, the shortest distance calculated in S22 is compared with a pre-set edge distance threshold. This threshold is used to define the range of the "transition zone." When the drone is close enough to the base station coverage boundary, it is considered to have entered the critical area where it is about to transition from a blind spot to a coverage area. When the shortest distance is less than this threshold, it is determined that the drone target has reached the edge transition zone of the base station coverage area, triggering the subsequent operations described in S2 (i.e., increasing the radio frequency tracking frequency, activating the cellular network monitoring module corresponding to the base station coverage area, and marking the pre-discovery task as pending association).

[0144] This embodiment pre-calibrates the signal coverage boundary coordinates of each cellular base station and establishes a base station coverage database, transforming the base station's engineering parameters into calculable coverage boundary coordinates, thus providing a data foundation for real-time distance calculation. During tracking, this database is invoked in real time to calculate the shortest distance between the UAV's real-time coordinates and the nearest signal coverage boundary, thereby accurately quantifying the UAV's current spatial margin relative to the cellular network coverage area. When this shortest distance is less than a preset edge distance threshold, it is determined that the UAV has reached the transition zone at the edge of the base station coverage area. The three stages sequentially execute a complete logical link of "pre-preparation, real-time calculation, and threshold determination," where the edge distance threshold can be flexibly configured according to different scenarios such as dense urban airspace or open suburban airspace, enabling the method to adapt to diverse geographical environments and monitoring needs.

[0145] In one embodiment, step S3 involves filtering candidate anonymous tracks from pre-discovery tasks marked as pending association, based on preset spatiotemporal correlation conditions. Specifically, this includes:

[0146] S31, based on the access time of the access terminal, sets a time window and filters out the pre-discovery tasks whose start time of the anonymous track falls within the time window.

[0147] Specifically, in this embodiment, a time window is set centered on the access time of the terminal access event captured in S3. All pre-discovery tasks marked as pending association are traversed, and the start time of the anonymous flight path recorded in each pre-discovery task is checked to see if it falls within this time window. If it does, it indicates that the UAV corresponding to the pre-discovery task is close to the access event in time, meeting the time consistency requirement, and proceeds to the next step of spatial filtering; otherwise, the pre-discovery task is excluded.

[0148] S32, extract the last location point of the UAV target in the selected pre-discovery task within the transition zone at the edge of the base station coverage area, and calculate the spatial distance between the last location point and the geographical coordinates of the cell where the access terminal is located.

[0149] Specifically, in this embodiment, for each pre-discovery task filtered by S31, the last location point of the UAV within the transition zone at the edge of the base station coverage area is extracted from its stored anonymous flight track; that is, the last location of the UAV before entering the coverage area by radio frequency positioning. Simultaneously, based on the cell-level positioning information obtained in S3, the geographic center coordinates of the cell to which the access terminal is connected are determined. Then, the spatial distance between the aforementioned last location point and the geographic coordinates of the cell is calculated. This distance reflects the spatial proximity between the UAV's last location before entering the coverage area and the terminal's network access location.

[0150] S33, when the spatial distance is less than the preset spatial distance threshold, the anonymous track corresponding to the pre-discovery task is listed as a candidate anonymous track.

[0151] Specifically, in this embodiment, the spatial distance calculated in S32 is compared with a preset spatial distance threshold. If the distance is less than the threshold, it indicates that the UAV corresponding to the pre-discovery task is spatially close to the location of the access terminal, satisfying the spatial consistency requirement. Pre-discovery tasks that simultaneously satisfy the temporal consistency in S31 and the spatial consistency in this step have their corresponding anonymous tracks listed as candidate anonymous tracks, entering the subsequent identity association verification stage.

[0152] This embodiment sets a time window based on the access time of the access terminal, filtering out pre-discovery tasks whose anonymous flight paths start times fall within this time window, thus completing the time-dimensional filtering. Next, it extracts the last location point of the UAV target within the transition zone at the edge of the base station coverage area from the filtered pre-discovery tasks, calculates the spatial distance between this location point and the geographic coordinates of the cell where the access terminal is located, thus completing the spatial distance calculation. Finally, it compares the calculated spatial distance with a preset spatial distance threshold. Only when the spatial distance is less than the threshold is the anonymous flight path corresponding to the pre-discovery task listed as a candidate anonymous flight path, thereby filtering out candidate targets that simultaneously meet both time and spatial constraints. The time window and spatial distance threshold can be flexibly configured according to actual scenarios such as UAV flight speed and base station deployment density, allowing the filtering accuracy to adapt to different environments. Simultaneously, the dual constraints significantly narrow down the candidate range that needs to enter subsequent hierarchical verification, effectively reducing computational complexity and the probability of false matching.

[0153] In one embodiment, step S4 further includes a third-level verification based on device identity information confirmation. This involves performing a fixed-order progressive hierarchical verification on the drone target corresponding to the candidate anonymous flight path and the access terminal to confirm identity association. Specifically, this includes:

[0154] S41, the first-level verification based on radio frequency fingerprint matching, includes:

[0155] S411, retrieve the radio frequency fingerprint corresponding to the candidate anonymous track.

[0156] Specifically, in this embodiment, the stored radio frequency fingerprint data is retrieved from the pre-discovery tasks associated with the anonymous tracks listed as candidates in S33, and used as the comparison benchmark for the first-level verification.

[0157] S412, Obtain the radio frequency capability field information reported by the access terminal to the cellular network.

[0158] Specifically, in this embodiment, the radio frequency capability field information (i.e., UE capability information) reported by the access terminal during the network access process is obtained from the cellular network side. This information includes the radio frequency capability parameters of the terminal's transmitter, such as transmit power level, modulation mode support capability, and frequency offset tolerance range. These parameters reflect the radio frequency characteristics of the terminal device's transmitter.

[0159] S413, a feature similarity algorithm is used to compare the radio frequency fingerprint with the radio frequency capability field information in multiple dimensions to obtain a comprehensive matching degree value. The comparison dimensions include at least the carrier frequency offset value and the phase noise spectrum feature.

[0160] Specifically, in this embodiment, the RF fingerprint retrieved in S411 and the RF capability field information obtained in S412 are compared in multiple dimensions. The comparison dimensions include at least two: carrier frequency offset value, comparing the degree of matching between the carrier frequency offset extracted from the RF fingerprint and the frequency offset tolerance range reflected in the RF capability field; and phase noise spectrum features, comparing the similarity between the shape of the phase noise spectrum in the RF fingerprint and the phase noise characteristics implicit in the RF capability field. A feature similarity algorithm is used to calculate a similarity score for each dimension, and then a weighted sum is obtained to obtain a comprehensive matching score.

[0161] S414. When the overall matching degree value is greater than the preset feature matching threshold, the first level of verification is passed and the second level of verification is entered. Otherwise, the identity association verification is failed and the current identity association verification process is terminated.

[0162] Specifically, in this embodiment, the comprehensive matching degree value obtained in S413 is compared with a preset feature matching threshold. If the comprehensive matching degree value is greater than the threshold, it indicates that the RF fingerprint and RF capability field are highly matched, and the first-level verification is passed. A medium-confidence association result is generated, indicating that the current evidence tends to suggest that the two belong to the same device, but it has not yet been finally confirmed. Then, the second-level verification is initiated. If the comprehensive matching degree value is not greater than the threshold, the identity association verification is directly determined to fail, and the entire verification process is terminated.

[0163] S42, the second-level verification based on kinematic properties, includes:

[0164] S421, calculate the spatial distance between the end point of the candidate anonymous track and the geographical coordinates of the cell where the access terminal is located. If the spatial distance is less than the preset spatial distance threshold, the spatial consistency verification is passed; otherwise, the identity association verification process is terminated.

[0165] Specifically, in this embodiment, the spatial distance between the end point of the candidate anonymous flight path and the geographic coordinates of the cell where the access terminal is located is calculated. This distance reflects the spatial proximity between the last perceived location of the UAV in the blind zone and the terminal's entry point into the network within the coverage area. If this distance is less than a preset threshold, it indicates that the two are spatially continuous, and the spatial consistency check is passed; otherwise, the check process is terminated.

[0166] S422, calculate the speed difference between the average speed of the UAV target corresponding to the candidate anonymous flight path during the radio frequency tracking phase and the instantaneous speed of the access terminal at the moment of network access. When the speed difference is less than the preset speed difference threshold, the speed continuity verification is passed; otherwise, the identity association verification process is terminated.

[0167] Specifically, in this embodiment, two speed values ​​are calculated: one is the average speed of the UAV target corresponding to the candidate anonymous flight path during the entire radio frequency tracking phase; the other is the instantaneous speed of the access terminal at the moment of network entry. The difference between the two is calculated. If the difference is less than a preset threshold, it indicates that there is no sudden change between the average flight speed of the UAV in the blind zone and the instantaneous speed at the moment of network entry, which meets the motion continuity requirement, and the speed continuity verification is passed; otherwise, the verification process is terminated.

[0168] S423, perform a smoothness check on the stitching of the candidate anonymous track and the real-name track of the access terminal within the cellular base station signal coverage area. If the rate of change of track curvature at the stitching point is less than the preset rate of change of curvature threshold, the track smoothness check is passed; otherwise, the identity association check process is terminated.

[0169] Specifically, in this embodiment, the end portion of the candidate anonymous track is simulated and spliced ​​with the beginning portion of the real-name track of the access terminal to check whether the track at the splicing point is smooth and continuous. The rate of change of track curvature at the splicing point is calculated, that is, the degree of change in track direction. If the rate of change of curvature is less than a preset threshold, it means that the two track segments can be naturally connected at the splicing point without any abnormalities such as jumps or sharp turns, and the track smoothness verification is passed; otherwise, the verification process is terminated.

[0170] S424: When the spatial consistency check, speed continuity check, and splicing smoothness check are all passed, the second level of check is passed, and the third level of check is entered.

[0171] Specifically, in this embodiment, the results of the three sub-checks S421, S422, and S423 are summarized. Only when all three checks pass is the second-level check considered passed and the third-level check proceeds. If any check fails, the entire check process is terminated.

[0172] S43, the third-level verification based on device identity information, includes:

[0173] S431, retrieve the historical archive of serial numbers of all unmanned aerial vehicles (UAVs).

[0174] Specifically, in this embodiment, a pre-established historical archive of UAV device serial numbers is retrieved. This archive stores the device serial numbers of all UAVs that have ever appeared within the coverage area of ​​this monitoring system and their corresponding historical records.

[0175] S432, retrieve whether the device serial number of the access terminal exists in the historical archive.

[0176] Specifically, in this embodiment, the device serial number of the access terminal is used as the keyword to search the historical archive retrieved in S431 to determine whether the serial number already has a filing record.

[0177] S433, if the device serial number of the access terminal exists in the historical archive, the access terminal is marked as a known device, and the identity of the UAV target corresponding to the candidate anonymous flight track is confirmed to be associated with the access terminal; if it does not exist, a new file is automatically created for the device serial number, and the identity of the UAV target corresponding to the candidate anonymous flight track is confirmed to be associated with the access terminal.

[0178] Specifically, in this embodiment, if the device serial number exists in the historical archive, the access terminal is marked as a known device, and the identity of the UAV target corresponding to the candidate anonymous flight path is finally confirmed to be associated with the access terminal, thus completing the identity association confirmation. If the device serial number does not exist in the historical archive, a new file record is automatically created for the device serial number, and then the identity association is confirmed in the same way. This marks the first time that the device has been identified and filed in this monitoring system.

[0179] This embodiment employs a fixed-order, three-layer progressive verification architecture to achieve identity association confirmation. The first layer performs hardware-level matching based on radio frequency fingerprints, using the unique physical characteristics caused by the nonlinearity of the transmitter hardware for preliminary screening, generating a medium-confidence association result upon successful screening. The second layer performs spatiotemporal consistency verification based on kinematic characteristics, sequentially performing spatial position deviation verification, velocity difference verification, and track curvature smoothness verification. Only when all three sub-verifications meet preset thresholds is the verification passed at this layer, and the confidence level is increased from medium to high confidence. The third layer confirms the device's identity information by searching the historical archive of UAV device serial numbers across the entire domain to complete the final confirmation of ownership. If the serial number already exists and is registered, it is marked as a known device; otherwise, a new file is automatically created. The three-layer verification forms a progressively converging verification chain of "hardware fingerprint anchoring, kinematic characteristic verification, and device file ownership confirmation." Each layer undertakes different verification responsibilities, progressing step by step and filtering out invalid matches. The second layer verifies motion continuity from three different dimensions: position, velocity, and trajectory morphology, effectively preventing erroneous associations caused by misjudgments in a single dimension. At the same time, it distinguishes between known devices and new devices through a historical archive, providing differentiated association confirmation methods for different types of devices, thus improving the completeness and adaptability of the method.

[0180] In one embodiment, step S423 involves verifying the smoothness of splicing between the candidate anonymous track and the real-name track of the access terminal within the cellular base station signal coverage area. This specifically includes:

[0181] Step a. Use spline interpolation algorithm to perform curve fitting between the N position points at the end of the candidate anonymous track and the M position points at the beginning of the real track, where N and M are both positive integers greater than or equal to 3.

[0182] Specifically, in this embodiment, N location points at the end of the candidate anonymous track and M location points at the beginning of the real track are selected. These two sets of location points are used as fitting data, and a spline interpolation algorithm is used to generate a smooth curve passing through all the fitting points. The spline interpolation algorithm can ensure that the curve has good continuity and smoothness at the splicing points. N and M are both positive integers greater than or equal to 3 to ensure that there are enough data points for effective curve fitting.

[0183] Step b. Calculate the rate of change of curvature of the fitted curve at the splice point.

[0184] Specifically, in this embodiment, the rate of change of curvature is calculated at the splicing point on the fitted curve. The rate of change of curvature reflects how quickly the curvature of the curve changes at that point. The smaller the rate of change of curvature, the smoother the curve at that point; the larger the rate of change of curvature, the sharper or abrupt the curve is at that point. This index is used to quantify the smoothness of the two track segments at the splicing point.

[0185] Step c. When the rate of curvature change is less than the preset rate of curvature change threshold, the trajectory smoothness condition is determined to be met.

[0186] Specifically, in this embodiment, the calculated rate of curvature change is compared with a preset rate of curvature change threshold. If the rate of curvature change is less than the threshold, it means that the end of the candidate anonymous track and the beginning of the real track can be naturally connected at the splicing point without any sudden change in direction or position, and the track smoothness condition is met, that is, the splicing smoothness verification of S423 is passed; if the rate of curvature change is not less than the threshold, the track smoothness condition is not met.

[0187] This embodiment employs a spline interpolation algorithm to perform curve fitting between N position points at the end of the candidate anonymous track and M position points at the beginning of the real track, where N and M are both positive integers greater than or equal to 3 to ensure sufficient data points to support effective fitting. After fitting, the rate of curvature change at the splicing point is calculated. This indicator can sensitively reflect the drastic change in track direction. Compared with directly comparing position deviation or heading angle, it can simultaneously capture the combined effect of position deviation and direction deviation, providing a more comprehensive assessment of track quality at the splicing point. When the rate of curvature change is less than a preset rate of curvature change threshold, the track smoothness condition is deemed met. The N points at the end of the candidate anonymous track represent the flight trajectory of the UAV in the last segment of the blind zone, and the M points at the beginning of the real track represent the flight trajectory of the UAV in the first segment after entering the network. If both belong to the same device, the two tracks should connect naturally at the splicing point, have the same direction, and have no abrupt changes. The spline interpolation algorithm can ensure that the fitted curve passes through all given location points and has second-order continuity, making the smoothness assessment more accurate. At the same time, the values ​​of N and M can be flexibly adjusted according to different sampling frequencies, avoiding fitting distortion caused by too few data points.

[0188] In one embodiment, step S5 involves cross-source fusion of the candidate anonymous flight track and the real-name flight track of the access terminal within the cellular base station signal coverage area, specifically including:

[0189] S51 identifies the differences between candidate anonymous tracks and real tracks in terms of coordinate reference and timestamp granularity.

[0190] Specifically, in this embodiment, the differences in format between the two types of track data are first identified. Candidate anonymous tracks originate from the radio frequency sensing side, and their location coordinates may be based on a local coordinate system, while their timestamps may be based on the device's local clock. Real-name tracks originate from the cellular network side, and their location coordinates are based on a global coordinate system, while their timestamps are based on the network's unified time. Identifying these differences is necessary to prepare for subsequent unified conversion.

[0191] S52 uses the moment of identity verification as the splicing point to uniformly transform the coordinates of candidate anonymous tracks and real tracks to the same coordinate system, thereby achieving spatial reference alignment.

[0192] Specifically, in this embodiment, the moment of identity association confirmation in S4 is used as the splicing point. The coordinates of the candidate anonymous track before the splicing point and the coordinates of the real track after the splicing point are uniformly converted to the same coordinate reference, so that the two types of tracks can be spliced ​​in the same reference system in space.

[0193] S53 unifies the timestamps of candidate anonymous tracks and real-name tracks to the same time base, achieving timeline alignment.

[0194] Specifically, in this embodiment, the timestamps of candidate anonymous tracks and real-name tracks are uniformly converted to the same time base, so that the two types of tracks can be correctly arranged in chronological order on the timeline.

[0195] S54 uses the Kalman filter algorithm to smooth the transition of the track at the splicing point.

[0196] Specifically, in this embodiment, a Kalman filter algorithm is used to smooth the trajectory near the splicing point. The Kalman filter establishes a motion state model, uses the trajectory data on both sides of the splicing point as observations, and estimates the optimal smooth trajectory at the splicing point, effectively eliminating trajectory jumps or discontinuities that may occur due to data source switching.

[0197] S55 binds and stores the stitched complete flight track with the device serial number of the access terminal.

[0198] Specifically, in this embodiment, the spliced ​​complete track is bound to the device serial number of the access terminal, so that the entire track has a real-name device identifier from the start to the end, and then stored in the track database for subsequent query and tracing.

[0199] This embodiment identifies the differences in coordinate reference and timestamp granularity between candidate anonymous tracks and real-name tracks, and transforms them to the same coordinate system and time reference, achieving spatial reference alignment and time axis alignment. This solves the heterogeneity problem in data format between radio frequency side tracks and cellular side tracks, enabling two types of data from different sources and with different formats to be stitched together under the same reference system. Using the moment of identity association confirmation as the stitching point, a Kalman filter algorithm is used to smooth the transition of the track at the stitching point. This algorithm can comprehensively consider the track data on both sides of the stitching point and generate the optimal smooth trajectory under the premise of satisfying kinematic constraints. This eliminates track jumps that may occur due to data source switching without losing effective information. After stitching, the complete track is bound and stored with the device serial number of the access terminal, so that the final output global track carries the real-name device identifier from the start point to the end point, achieving full identity continuity from the blind spot to the coverage area.

[0200] In one embodiment, the radio frequency and cellular collaborative networked drone full-domain monitoring method further includes:

[0201] S6 When multiple drone targets arrive at the edge transition zone of the base station coverage area at the same time and there are multiple terminal access events, the preset multi-target anti-confusion processing strategy is executed.

[0202] Specifically, in this embodiment, when multiple drone targets arrive at the edge of the base station coverage area almost simultaneously, and these drones initiate cellular network access almost simultaneously after entering the coverage area, multiple pre-discovery tasks and multiple terminal access events will be generated, and they may be very close in time and space. In this situation, if the single-target association process is followed, there is a risk that a pre-discovery task may be incorrectly associated with a certain access terminal. Therefore, a preset multi-target anti-confusion processing strategy is implemented to ensure that each pre-discovery task is correctly associated with the corresponding access terminal.

[0203] This embodiment adds specialized processing for multi-target scenarios to the single-target association process. When multiple pre-discovery tasks are detected simultaneously in the pending association state and the cellular network monitoring module simultaneously captures multiple terminal access events, a multi-target anti-confusion processing mode is entered, and a preset processing strategy is executed. This strategy is a customized anti-confusion mechanism for dense multi-target scenarios, solving the identity mismatch problem that is prone to occur when multiple drones enter and exit the coverage area at the same time. At the same time, this processing step can be executed before cross-source fusion, ensuring that the optimal matching in multi-target scenarios is fully considered in the identity association confirmation stage, avoiding the extra overhead of reprocessing after discovering incorrect matching after completing single-target association.

[0204] In one embodiment, step S6 includes the following steps in the preset multi-target anti-obfuscation processing strategy:

[0205] S61, construct the correlation scoring matrix, where the matrix rows correspond to each candidate anonymous track and the matrix columns correspond to each access terminal.

[0206] Specifically, in this embodiment, a two-dimensional scoring matrix is ​​constructed based on all candidate anonymous tracks and all access terminals to be associated. Each row of the matrix corresponds to a candidate anonymous track, each column corresponds to an access terminal, and each cell of the matrix represents the matching score between the candidate anonymous track in that row and the access terminal in that column.

[0207] S62 calculates the comprehensive matching score for each candidate anonymous track-access terminal pair based on radio frequency fingerprint similarity and kinematic matching degree.

[0208] Specifically, in this embodiment, for each cell in the scoring matrix, i.e., each pair of candidate anonymous tracks and access terminals, a matching score across two dimensions is calculated. The first dimension is radio frequency fingerprint similarity, which is the degree of similarity between the radio frequency fingerprint of the candidate anonymous track and the radio frequency capability field of the access terminal. The second dimension is kinematic matching, which is the degree of matching of kinematic features such as spatial proximity, speed consistency, and track smoothness between the terminal location of the candidate anonymous track and the geographical coordinates of the cell where the access terminal is located. The scores of the two dimensions are then weighted and summed according to preset weights to obtain the comprehensive matching score for that pair.

[0209] S63, select the candidate anonymous track-access terminal pair with the highest global score and a comprehensive matching score not lower than the preset matching threshold to complete the one-to-one identity binding, and remove the bound candidate anonymous track and access terminal from the matrix.

[0210] Specifically, in this embodiment, the candidate anonymous track and access terminal pair with the highest comprehensive matching score are searched in the scoring matrix. If the highest score is not lower than a preset matching threshold, the pair is identity-bound to confirm that the candidate anonymous track and the access terminal belong to the same UAV. After binding, the row containing the track and the column containing the terminal are removed from the matrix to prevent the track or terminal from being matched again.

[0211] S64, repeat steps S62 to S63 until all candidate anonymous tracks and access terminals in the matrix have been bound or there are no matching objects.

[0212] Specifically, in this embodiment, operations S62 and S63 are repeated: at the beginning of each loop, the comprehensive matching score for the remaining candidate anonymous tracks and access terminals in the matrix is ​​recalculated, and then the highest global score in the remaining matrix is ​​selected for binding and eliminated. The loop continues until all candidate anonymous tracks and access terminals in the matrix have been bound, or there are no candidate anonymous tracks and access terminals in the matrix with a score not lower than the preset matching threshold.

[0213] S65 marks candidate anonymous track-access terminal pairs whose comprehensive matching scores are lower than the preset matching threshold as pending manual review.

[0214] Specifically, in this embodiment, after the loop ends, there may still be some candidate anonymous tracks and access terminal pairs in the matrix whose overall matching scores are lower than the preset matching threshold and cannot be automatically bound. These low-scoring pairs are marked as pending manual review, and further analysis and confirmation are performed manually to avoid missing real associations due to overly strict automatic matching thresholds.

[0215] This embodiment constructs an association scoring matrix to quantify the matching relationships between all candidate anonymous tracks and all access terminals into a unified scoring framework, transforming the matching problem in multi-target scenarios into a computable optimization problem. Based on two dimensions—RF fingerprint similarity and kinematic matching degree—the comprehensive matching score of each candidate anonymous track and access terminal pair is calculated, ensuring that the matching confidence in multi-target scenarios maintains the same judgment standard as in single-target scenarios. Each time, the pair with the highest comprehensive matching score in the remaining matrix, which is not lower than a preset threshold, is selected to complete a one-to-one identity binding, and the bound tracks and terminals are removed from the matrix, ensuring that in cases where multiple targets are competing for the same access terminal... The highest-scoring match for the same flight path is prioritized for confirmation. The above selection and binding operations are repeated until all candidate anonymous flight paths and access terminals in the matrix are bound or there are no matching objects. Regardless of the number of drones entering the coverage area at the same time, they can be correctly associated one by one, adapting to the large-scale monitoring needs in high-density flight scenarios. Finally, candidate anonymous flight paths and access terminal pairs with a comprehensive matching score lower than the preset matching threshold are marked as awaiting manual review. This provides a fallback mechanism for boundary cases that cannot be covered by automatic matching, avoiding missing real associations due to overly strict thresholds or forced mismatches due to low confidence, further improving the reliability of matching and the fault tolerance of the system.

[0216] like Figure 2 As shown in the figure, this application provides a networked drone all-domain monitoring system that combines radio frequency and cellular collaboration. The system may include:

[0217] The radio frequency sensing module 201 is deployed in areas not covered by cellular base station signals. It is configured to perform passive sensing of UAV targets, acquire the radio frequency fingerprint and location information of UAV targets and continuously track them, form anonymous tracks, and create corresponding pre-discovery tasks based on the anonymous tracks.

[0218] The transition detection module 202 is configured to monitor the position of the UAV target corresponding to the pre-discovery task. When it is determined that the UAV target has reached the transition zone at the edge of the base station coverage area, the radio frequency tracking frequency is increased, the cellular network monitoring module corresponding to the base station coverage area is activated, and the pre-discovery task is marked as pending association.

[0219] Cellular network monitoring module 203 is configured to capture terminal access events within the base station coverage area and obtain the identity information and access spatiotemporal information of the accessing terminal.

[0220] The track filtering module 204 is configured to filter candidate anonymous tracks from pre-discovery tasks marked as pending association based on preset spatiotemporal association conditions.

[0221] The identity association verification module 205 is configured to perform a fixed-order progressive hierarchical verification on the UAV target corresponding to the candidate anonymous flight path and the access terminal to confirm the identity association. The progressive hierarchical verification includes at least a first-level verification based on radio frequency fingerprint matching and a second-level verification based on kinematic characteristics verification.

[0222] The cross-source fusion output module 206 is configured to, in response to identity association confirmation, perform cross-source fusion of candidate anonymous flight tracks and real-name flight tracks of access terminals within the cellular base station signal coverage area, and output a full-domain continuous flight track with access terminal identity information.

[0223] In one embodiment, the radio frequency sensing module 201 includes:

[0224] A distributed radio frequency sensor array deployed in areas not covered by cellular base station signals, with the hardware and data link of the radio frequency sensor array independent of the cellular network;

[0225] The radio frequency sensor array is configured to: capture UAV radio frequency signals, extract physical layer features including carrier frequency offset, phase noise, and protocol type from the UAV radio frequency signals, and construct a unique radio frequency fingerprint of the UAV target; calculate the real-time coordinates of the UAV target using a time difference of arrival (TDOA) or angle of arrival (AGA) positioning algorithm as location information; assign a unique radio frequency identifier to the UAV target, and create a corresponding pre-discovery task based on the unique radio frequency identifier, storing the radio frequency fingerprint, location information, and discovery time in the pre-discovery task; continuously collect the UAV target's location data at a fixed sampling period, and accumulate it into a continuous anonymous flight track according to the time series, storing it in the pre-discovery task library.

[0226] In one embodiment, the transition detection module 202 is specifically configured as follows:

[0227] The signal coverage boundary coordinates of each cellular base station are pre-calibrated to establish a base station coverage database. The base station coverage database is called to calculate the shortest distance between the real-time coordinates of the UAV target and the signal coverage boundary closest to the real-time coordinates of the UAV target. When the shortest distance is less than the preset edge distance threshold, it is determined that the UAV target has reached the edge transition zone of the base station coverage area and the radio frequency tracking frequency is increased.

[0228] In one embodiment, the track filtering module 204 is specifically configured as follows:

[0229] Based on the access time of the access terminal, a time window is set to filter out pre-discovery tasks whose start time of anonymous tracks falls within the time window; the last location point of the UAV target in the selected pre-discovery task within the transition zone at the edge of the base station coverage area is extracted, and the spatial distance between the last location point and the geographical coordinates of the cell where the access terminal is located is calculated; when the spatial distance is less than the preset spatial distance threshold, the anonymous track corresponding to the pre-discovery task is listed as a candidate anonymous track.

[0230] In one embodiment, the identity association verification module 205 is specifically configured as follows:

[0231] The first-level verification based on radio frequency fingerprint matching is performed, including: retrieving the radio frequency fingerprint corresponding to the candidate anonymous track; obtaining the radio frequency capability field information reported by the access terminal to the cellular network; using a feature similarity algorithm to compare the radio frequency fingerprint with the radio frequency capability field information in multiple dimensions to obtain a comprehensive matching degree value, wherein the comparison dimensions include at least the carrier frequency offset value and the phase noise spectrum feature; when the comprehensive matching degree value is greater than the preset feature matching threshold, the first-level verification is determined to be passed and the second-level verification is entered; otherwise, the identity association verification is determined to be failed and the current identity association verification process is terminated.

[0232] The second-level verification based on kinematic characteristics is performed, including: calculating the spatial distance between the end point of the candidate anonymous track and the geographic coordinates of the cell where the access terminal is located; if the spatial distance is less than a preset spatial distance threshold, the spatial consistency verification is passed; calculating the speed difference between the average speed of the UAV target corresponding to the candidate anonymous track during the radio frequency tracking phase and the instantaneous speed of the access terminal at the moment of network access; if the speed difference is less than a preset speed difference threshold, the speed continuity verification is passed; performing a stitching smoothness verification on the candidate anonymous track and the real-name track of the access terminal within the cellular base station signal coverage area; if the rate of change of track curvature at the stitching point is less than a preset rate of change of curvature threshold, the track smoothness verification is passed; if the spatial consistency verification, speed continuity verification, and stitching smoothness verification are all passed, the second-level verification is passed, and the process proceeds to the third-level verification.

[0233] Perform a third-level verification based on device identity information, including: retrieving the historical archive of UAV device serial numbers across the entire domain; checking whether the access terminal's device serial number exists in the historical archive; if the access terminal's device serial number exists in the historical archive, then mark the access terminal as a known device and confirm that the UAV target corresponding to the candidate anonymous flight path is associated with the access terminal's identity; if it does not exist, then automatically create a new file for the device serial number and confirm that the UAV target corresponding to the candidate anonymous flight path is associated with the access terminal's identity.

[0234] In one embodiment, the cross-source fusion output module 206 is specifically configured as follows:

[0235] The system identifies the differences between candidate anonymous flight tracks and real-name flight tracks in terms of coordinate reference and timestamp granularity; using the moment of identity confirmation as the splicing point, it unifies the coordinates of candidate anonymous flight tracks and real-name flight tracks to the same coordinate system to achieve spatial reference alignment; it unifies the timestamps of candidate anonymous flight tracks and real-name flight tracks to the same time reference to achieve time axis alignment; it uses a Kalman filter algorithm to perform smooth transition processing on the flight tracks at the splicing point; it binds and stores the spliced ​​complete flight track with the device serial number of the access terminal, and outputs a full-domain continuous flight track with access terminal identity information.

[0236] In one embodiment, the identity association verification module 205 is further configured as follows:

[0237] When multiple drone targets simultaneously arrive at the edge of the base station coverage area and correspond to multiple terminal access events, a preset multi-target anti-aliasing processing strategy is executed, including:

[0238] Construct an association scoring matrix, with rows corresponding to each candidate anonymous track and columns corresponding to each access terminal. Calculate the comprehensive matching score for each candidate anonymous track-access terminal pair based on RF fingerprint similarity and kinematic matching degree. Select the candidate anonymous track-access terminal pair with the highest global score and a comprehensive matching score not lower than a preset matching threshold to complete one-to-one identity binding, and remove the bound candidate anonymous tracks and access terminals from the matrix. Repeat the above selection and binding operations until all candidate anonymous tracks and access terminals in the matrix are bound or there are no matching objects. Mark candidate anonymous track-access terminal pairs with comprehensive matching scores lower than the preset matching threshold as awaiting manual review.

[0239] It should be noted that the radio frequency and cellular collaborative networked UAV all-domain monitoring system in the above embodiments has the same working principle and technical effect as the radio frequency and cellular collaborative networked UAV all-domain monitoring method in the above embodiments, and will not be repeated here.

[0240] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the apparatus disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the description is relatively simple; relevant parts can be referred to the method section.

[0241] Those skilled in the art will further recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0242] The steps of the methods or algorithms described in conjunction with the embodiments disclosed herein can be implemented directly by hardware, a software module executed by a processor, or a combination of both. The software module can be located in random access memory (RAM), main memory, read-only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disk, removable disk, CD-ROM, or any other form of storage medium known in the art.

[0243] The above description of the disclosed embodiments enables those skilled in the art to make or use this application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this application. Therefore, this application is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for full-domain monitoring of networked unmanned aerial vehicles (UAVs) using a combination of radio frequency and cellular technologies, characterized in that, The method includes the following steps: S1, In areas not covered by cellular base station signals, passive sensing is performed on the UAV target through radio frequency sensing equipment to obtain the radio frequency fingerprint and location information of the UAV target and continuously track it to form an anonymous flight path, and a corresponding pre-discovery task is created based on the anonymous flight path; S2, monitor the location of the drone target corresponding to the pre-discovery task; when it is determined that the drone target has reached the transition zone at the edge of the base station coverage area, increase the radio frequency tracking frequency, start the cellular network monitoring module corresponding to the base station coverage area, and mark the pre-discovery task as pending association. S3, when the cellular network monitoring module captures a terminal access event within the base station coverage area, it obtains the identity information and access spatiotemporal information of the access terminal, and selects candidate anonymous tracks from the pre-discovery tasks marked as pending association based on preset spatiotemporal association conditions. S4, perform a fixed-order progressive hierarchical verification between the UAV target corresponding to the candidate anonymous flight path and the access terminal to confirm the identity association, wherein the progressive hierarchical verification includes at least a first-level verification based on radio frequency fingerprint matching and a second-level verification based on kinematic characteristics verification. S5, in response to identity association confirmation, the candidate anonymous flight track is fused with the real-name flight track of the access terminal in the cellular base station signal coverage area, and the full-domain continuous flight track with the identity information of the access terminal is output.

2. The method for full-domain monitoring of networked drones using radio frequency and cellular coordination according to claim 1, characterized in that, S1 specifically includes: S11, a radio frequency sensor array is deployed in a distributed manner in the area not covered by the cellular base station signal, the hardware and data link of the radio frequency sensor array being independent of the cellular network; S12, capture the UAV radio frequency signal through the radio frequency sensor array, extract the physical layer features of the UAV radio frequency signal including carrier frequency offset, phase noise and protocol type, and construct a unique radio frequency fingerprint of the UAV target; S13, The real-time coordinates of the UAV target are calculated using a time difference of arrival (TDOA) positioning algorithm or an angle of arrival (AGA) positioning algorithm, and used as the location information; S14, assign a unique radio frequency (RF) identifier to the drone target, and create a corresponding pre-discovery task based on the RF identifier, and store the RF fingerprint, the location information and the discovery time in the pre-discovery task; S15, continuously collect the position data of the UAV target at a fixed sampling period, and accumulate them into a continuous anonymous flight track according to the time series, and store it in the pre-discovery task library.

3. The method for full-domain monitoring of networked drones using radio frequency and cellular coordination as described in claim 1, characterized in that, In step S2, determining that the drone target has reached the transition zone at the edge of the base station coverage area specifically includes: S21, Pre-calibrate the signal coverage boundary coordinates of each cellular base station and establish a base station coverage database; S22, call the base station coverage database to calculate the shortest distance between the real-time coordinates of the UAV target and the signal coverage boundary closest to the real-time coordinates of the UAV target; S23, when the shortest distance is less than the preset edge distance threshold, it is determined that the drone target has reached the edge transition zone of the base station coverage area.

4. The method for full-domain monitoring of networked drones using radio frequency and cellular coordination according to claim 1, characterized in that, In step S3, the step of selecting candidate anonymous tracks from the pre-discovery tasks marked as pending association, based on preset spatiotemporal correlation conditions, specifically includes: S31, based on the access time of the access terminal, set a time window and filter out the pre-discovery tasks whose start time of the anonymous track falls within the time window; S32, extract the last location point of the UAV target in the filtered pre-discovery task within the transition zone at the edge of the base station coverage area, and calculate the spatial distance between the last location point and the geographical coordinates of the cell where the access terminal is located; S33, when the spatial distance is less than a preset spatial distance threshold, the anonymous track corresponding to the pre-discovery task is listed as a candidate anonymous track.

5. The method for full-domain monitoring of networked drones using radio frequency and cellular coordination according to claim 1, characterized in that, In step S4, the progressive hierarchical verification further includes a third-level verification based on device identity information confirmation. Specifically, performing a fixed-order progressive hierarchical verification on the UAV target corresponding to the candidate anonymous flight path and the access terminal to confirm identity association includes: S41, the first-level verification based on radio frequency fingerprint matching, includes: S411, retrieve the radio frequency fingerprint corresponding to the candidate anonymous track; S412, Obtain the radio frequency capability field information reported by the access terminal to the cellular network; S413, The feature similarity algorithm is used to compare the radio frequency fingerprint with the radio frequency capability field information in multiple dimensions to obtain a comprehensive matching degree value. The comparison dimensions include at least the carrier frequency offset value and the phase noise spectrum feature. S414, when the comprehensive matching degree value is greater than the preset feature matching threshold, it is determined that the first level of verification has passed and the second level of verification has been entered; otherwise, it is determined that the identity association verification has failed and the current identity association verification process is terminated. S42, the second-level verification based on kinematic properties, includes: S421, calculate the spatial distance between the end point of the candidate anonymous track and the geographical coordinates of the cell where the access terminal is located. When the spatial distance is less than the preset spatial distance threshold, the spatial consistency verification is passed; otherwise, the identity association verification process is terminated. S422, calculate the speed difference between the average speed of the UAV target corresponding to the candidate anonymous flight path during the radio frequency tracking phase and the instantaneous speed of the access terminal at the moment of network access. When the speed difference is less than the preset speed difference threshold, it is determined that the speed continuity verification is passed; otherwise, the identity association verification process is terminated. S423, perform splicing smoothness verification on the candidate anonymous track and the real-name track of the access terminal in the cellular base station signal coverage area. When the track curvature change rate at the splicing point is less than the preset curvature change rate threshold, the track smoothness verification is passed; otherwise, the identity association verification process is terminated. S424, when all the spatial consistency check, speed continuity check and splicing smoothness check are passed, the second level of check is passed and the third level of check is entered; S43, the third-level verification based on device identity information, includes: S431, retrieve the historical archive of serial numbers of all unmanned aerial vehicles (UAVs) in the entire domain; S432, retrieve whether the device serial number of the access terminal exists in the historical archive; S433, if the device serial number of the access terminal exists in the historical archive, then the access terminal is marked as a known device, and the drone target corresponding to the candidate anonymous flight path is confirmed to be associated with the access terminal's identity; if it does not exist, then a new file is automatically created for the device serial number, and the drone target corresponding to the candidate anonymous flight path is confirmed to be associated with the access terminal's identity.

6. The method for full-domain monitoring of networked unmanned aerial vehicles (UAVs) using radio frequency and cellular coordination according to claim 5, characterized in that, In step S423, the step of performing a smoothness check on the stitching of the candidate anonymous flight track and the real-name flight track of the access terminal within the cellular base station signal coverage area specifically includes: A spline interpolation algorithm is used to perform curve fitting between the N position points at the end of the candidate anonymous track and the M position points at the beginning of the real track, where N and M are both positive integers greater than or equal to 3; Calculate the rate of change of curvature of the fitted curve at the splice point; When the rate of curvature change is less than a preset rate of curvature change threshold, the trajectory smoothness condition is determined to be met.

7. The method for full-domain monitoring of networked unmanned aerial vehicles (UAVs) using radio frequency and cellular coordination according to claim 1, characterized in that, In step S5, the cross-source fusion of the candidate anonymous flight track and the real-name flight track of the access terminal within the cellular base station signal coverage area specifically includes: S51, identify the differences between the candidate anonymous track and the real track in terms of coordinate reference and timestamp granularity; S52, using the moment of identity association confirmation as the splicing point, the coordinates of the candidate anonymous track and the coordinates of the real track are uniformly transformed to the same coordinate system to achieve spatial reference alignment; S53, unify the timestamps of the candidate anonymous tracks and the timestamps of the real-name tracks to the same time base to achieve timeline alignment; S54 uses the Kalman filter algorithm to smooth the transition of the track at the splicing point; S55, the spliced ​​complete flight track is bound and stored with the device serial number of the access terminal.

8. The radio frequency and cellular collaborative networked UAV full-domain monitoring method according to any one of claims 1-7, characterized in that, Also includes: S6 When multiple drone targets arrive at the edge of the base station coverage area at the same time and there are multiple terminal access events, the preset multi-target anti-confusion processing strategy is executed.

9. The method for full-domain monitoring of networked unmanned aerial vehicles using radio frequency and cellular coordination according to claim 8, characterized in that, In step S6, the preset multi-target anti-obfuscation processing strategy includes the following steps: S61, construct the correlation scoring matrix, where the matrix rows correspond to each candidate anonymous track and the matrix columns correspond to each access terminal; S62, calculate the comprehensive matching score for each candidate anonymous track-access terminal pair based on radio frequency fingerprint similarity and kinematic matching degree; S63, select the candidate anonymous track-access terminal pair with the highest global score and the comprehensive matching score not lower than the preset matching threshold to complete the one-to-one identity binding, and remove the bound candidate anonymous track and access terminal from the matrix; S64, repeat steps S62 to S63 until all candidate anonymous tracks and access terminals in the matrix have been bound or there are no matching objects; S65 marks candidate anonymous track-access terminal pairs whose comprehensive matching scores are lower than the preset matching threshold as pending manual review.

10. A networked unmanned aerial vehicle (UAV) all-domain monitoring system that combines radio frequency and cellular technologies, characterized in that: The system includes: The radio frequency sensing module is deployed in areas not covered by cellular base station signals. It is configured to perform passive sensing of UAV targets, acquire the radio frequency fingerprint and location information of the UAV targets and continuously track them to form anonymous tracks, and create corresponding pre-discovery tasks based on the anonymous tracks. The transition detection module is configured to monitor the location of the UAV target corresponding to the pre-discovery task. When it is determined that the UAV target has reached the transition zone at the edge of the base station coverage area, the radio frequency tracking frequency is increased, the cellular network monitoring module corresponding to the base station coverage area is activated, and the pre-discovery task is marked as pending association. The cellular network monitoring module is configured to capture terminal access events within the coverage area of ​​the base station and obtain the identity information and spatiotemporal information of the accessing terminal. The track filtering module is configured to filter candidate anonymous tracks from pre-discovery tasks marked as pending association based on preset spatiotemporal association conditions. The identity association verification module is configured to perform a fixed-order progressive hierarchical verification between the UAV target corresponding to the candidate anonymous flight path and the access terminal to confirm the identity association. The progressive hierarchical verification includes at least a first-level verification based on radio frequency fingerprint matching and a second-level verification based on kinematic characteristics verification. The cross-source fusion output module is configured to, in response to identity association confirmation, perform cross-source fusion of the candidate anonymous flight track and the real-name flight track of the access terminal within the cellular base station signal coverage area, and output a full-domain continuous flight track with the identity information of the access terminal.

Images