CN120408226A

CN120408226A - A method and system for quickly identifying specific drone models based on radio frequency fingerprint library

Info

Publication number: CN120408226A
Application number: CN202510913686.1A
Authority: CN
Inventors: 娄山良; 杨连臣; 孙亚宾; 李小彤; 杨谕黔; 方厚章
Original assignee: Zhonglian Deguan Technology Beijing Co ltd
Current assignee: Zhonglian Deguan Technology Beijing Co ltd
Priority date: 2025-07-03
Filing date: 2025-07-03
Publication date: 2025-08-01
Anticipated expiration: 2045-07-03
Also published as: CN120408226B

Abstract

The application provides a method and a system for quickly identifying a specific unmanned aerial vehicle model based on a radio frequency fingerprint library. The method comprises the steps of firstly receiving radio frequency signals generated by a target unmanned aerial vehicle in flight, dynamically denoising the radio frequency signals to generate denoised radio frequency signals, separating combined data containing a signal intensity fluctuation mode, a phase jump frequency and modulation distortion characteristics from the denoised radio frequency signals, performing layer-by-layer matching on the combined data and a pre-built radio frequency fingerprint library to generate a matching result, extracting model association parameters of the target unmanned aerial vehicle from the matching result, generating a recognition decision containing a real-time environment adaptation factor based on the model association parameters, and outputting a target unmanned aerial vehicle model number result corresponding to the recognition decision.

Description

Method and system for quickly identifying specific unmanned aerial vehicle model based on radio frequency fingerprint library

Technical Field

The application relates to the technical field of unmanned aerial vehicles, in particular to a method and a system for quickly identifying a specific unmanned aerial vehicle model based on a radio frequency fingerprint library.

Background

Along with the wide application of unmanned aerial vehicle in fields such as security protection monitoring, logistics transportation, etc., aiming at complex electromagnetic interference and multi-obstacle reflection open scenes (such as urban building groups, industrial factories, etc.), the realization of quick and accurate identification of specific unmanned aerial vehicle models is needed. In such a scene, the radio frequency signals of the unmanned aerial vehicle are easy to be interfered by the environment to cause characteristic distortion, and the signal difference of unmanned aerial vehicles of different types is fine, so that signal processing and model matching are required to be completed within millisecond response time, and the real-time supervision and countercheck requirements are met.

At present, a matching recognition scheme based on a static radio frequency fingerprint library is mainly adopted for the requirements. According to the scheme, radio frequency signals of unmanned aerial vehicles of different types in a fixed experimental environment are pre-collected, signal intensity, spectrum characteristics and the like are extracted to generate a reference fingerprint library, and when in actual identification, a real-time signal of a target unmanned aerial vehicle is subjected to threshold comparison with the fingerprint library, and a model with the highest matching degree is screened to be used as an identification result.

The scheme has obvious limitations in a dynamic open scene that firstly, a static fingerprint library cannot be adapted to electromagnetic interference intensity and an obstacle reflection path which change in real time in a complex environment, so that the offset of signal characteristics and a reference fingerprint exceeds a preset threshold value to cause misjudgment, secondly, a fixed threshold matching mechanism is difficult to cope with dynamic fluctuation of signal distortion degree, the threshold value is required to be adjusted for multiple times or the fingerprint library is recalibrated, the recognition instantaneity is seriously affected, and thirdly, a collaborative mechanism of environment parameters (such as interference source distribution and obstacle positions) and radio frequency fingerprint correction is not established, so that the recognition efficiency difference in different scenes is extremely large.

Disclosure of Invention

The application provides a method and a system for quickly identifying a specific unmanned aerial vehicle model based on a radio frequency fingerprint library, which are used for solving the problems that in the prior art, a static radio frequency fingerprint library cannot be adapted to signal offset caused by electromagnetic interference and obstacle reflection of a dynamic open scene, the real-time performance of a fixed threshold matching mechanism caused by environmental fluctuation is insufficient, and the identification efficiency is unstable caused by lack of cooperative association of environmental parameters and radio frequency fingerprint correction.

In a first aspect, the application provides a method for quickly identifying a specific unmanned aerial vehicle model based on a radio frequency fingerprint library, which comprises the following steps:

Receiving a radio frequency signal generated by a target unmanned aerial vehicle in flight, and carrying out dynamic noise reduction treatment on the radio frequency signal to generate a noise-reduced radio frequency signal;

Separating combined data containing a signal intensity fluctuation mode, a phase jump frequency and a modulation distortion characteristic from the radio frequency signal after noise reduction, and carrying out layer-by-layer matching on the combined data and a pre-constructed radio frequency fingerprint library to generate a matching result;

Extracting model-related parameters of the target unmanned aerial vehicle from the matching result, wherein the model-related parameters comprise signal propagation delay differences corresponding to specific obstacle reflection paths and frequency domain distortion association degrees corresponding to electromagnetic interference source distribution;

And generating an identification decision containing a real-time environment adaptation factor based on the model-related parameters, and outputting a target unmanned aerial vehicle model number result corresponding to the identification decision.

Optionally, the method is characterized by receiving a radio frequency signal generated by the target unmanned aerial vehicle in flight, performing dynamic noise reduction processing on the radio frequency signal, and generating a noise-reduced radio frequency signal, and includes:

capturing an original radio frequency signal of a target unmanned aerial vehicle in real time through multichannel signal acquisition equipment deployed in a monitoring area, wherein the original radio frequency signal comprises a noise component caused by electromagnetic interference sources and obstacle reflections in a current scene;

monitoring environmental parameters in the original radio frequency signal, wherein the environmental parameters comprise real-time interference intensity distribution and obstacle density distribution;

selecting a signal filtering intensity level according to the environment parameters, performing layer-by-layer noise suppression on the original radio frequency signal by adopting multistage series filtering if the real-time interference intensity is higher than a set range and the obstacle density is higher than a set threshold, and performing noise suppression on the original radio frequency signal by adopting single-stage filtering if the real-time interference intensity is lower than the set range to obtain a filtered radio frequency signal;

And carrying out signal integrity recovery processing on the filtered radio frequency signals to generate noise-reduced radio frequency signals.

Optionally, separating combined data including a signal intensity fluctuation mode, a phase jump frequency and a modulation distortion characteristic from the radio frequency signal after noise reduction, and performing layer-by-layer matching on the combined data and a pre-constructed radio frequency fingerprint library to generate a matching result, wherein the method comprises the following steps of:

dividing the noise-reduced radio frequency signal into a plurality of signal segments according to a preset time window, and extracting the fluctuation mode of the signal intensity along with the change of the flying height, the frequency distribution of the phase jump interval and the position characteristics of modulation waveform distortion from each signal segment;

Fusing the fluctuation mode, the frequency distribution and the position characteristics into combined data according to a preset weight proportion;

sequentially carrying out similarity calculation on the combined data and a first layer signal intensity fluctuation mode reference set stored in a radio frequency fingerprint library, and screening a candidate model set with similarity higher than a first threshold value;

And carrying out joint matching on the frequency distribution and the position characteristics in the combined data and the second-layer phase jump frequency reference set and the modulation distortion reference set corresponding to the candidate model set to generate a matching result containing the matching degree ordering of the candidate models.

Optionally, extracting a model-related parameter of the target unmanned aerial vehicle from the matching result, where the model-related parameter includes a signal propagation delay difference corresponding to a specific obstacle reflection path and a frequency domain distortion association degree corresponding to an electromagnetic interference source distribution, and the method includes:

According to the matching degree ranking of the candidate models in the matching result, extracting radio frequency fingerprint reference data corresponding to the highest candidate model;

acquiring signal propagation delay difference parameters related to obstacle distribution in a current scene from the radio frequency fingerprint reference data;

acquiring frequency domain distortion association degree parameters associated with electromagnetic interference source distribution from the radio frequency fingerprint reference data;

and integrating the signal propagation delay difference parameter and the frequency domain distortion correlation parameter into a model correlation parameter of the target unmanned aerial vehicle.

Optionally, generating an identification decision including a real-time environment adaptation factor based on the model-related parameter, and outputting a target unmanned model number result corresponding to the identification decision, including:

acquiring real-time tracking data of dynamic changes of the position of an obstacle in a current scene and real-time fluctuation data of the intensity of an electromagnetic interference source;

Adjusting the weight coefficient of the signal propagation delay difference parameter according to the real-time tracking data, and adjusting the weight coefficient of the frequency domain distortion association parameter according to the real-time fluctuation data;

Inputting the adjusted signal propagation delay difference parameters and the adjusted frequency domain distortion association parameters into an environment adaptation factor calculation model to generate real-time environment adaptation factors for correcting the matching degree of candidate models in the radio frequency fingerprint library;

dynamically compensating the matching degree of the candidate model in the matching result according to the real-time environment adaptation factor, and generating an identification decision containing the final matching model;

and carrying out consistency verification on the model identification of the final matching model in the identification decision and the model feature description prestored in the radio frequency fingerprint library, and outputting the final matching model as a target unmanned aerial vehicle model result when the consistency verification passes.

Optionally, adjusting a weight coefficient of a signal propagation delay difference parameter according to the real-time tracking data, and adjusting a weight coefficient of a frequency domain distortion association parameter according to the real-time fluctuation data, including:

Dynamically adjusting the weight coefficient of the signal propagation delay difference parameter according to the dynamic rate of the position change of the obstacle in the real-time tracking data, and linearly enhancing the weight coefficient according to the ratio of the dynamic rate to the preset rate threshold when the dynamic rate exceeds the preset rate threshold to generate an adjusted delay weight coefficient;

Dynamically adjusting the weight coefficient of the frequency domain distortion association parameter according to the fluctuation amplitude of the electromagnetic interference source intensity in the real-time fluctuation data, and when the fluctuation amplitude exceeds a preset amplitude threshold value, gradually increasing the weight coefficient according to the difference value of the fluctuation amplitude and the amplitude threshold value to generate an adjusted frequency domain weight coefficient.

Optionally, dynamically compensating the matching degree of the candidate model in the matching result according to the real-time environment adaptation factor, and generating an identification decision including the final matching model, including:

decomposing the real-time environment adaptation factor into a time delay correction factor and a frequency domain correction factor;

Extracting initial matching degree of the candidate model set from the matching result, and multiplying the time delay correction factor by a signal propagation time delay difference parameter corresponding to each candidate model in the initial matching degree item by item to generate a time delay compensated matching degree;

Carrying out item-by-item multiplication on the frequency domain correction factors and frequency domain distortion association parameters corresponding to each candidate model in the time delay compensated matching degree to generate the frequency domain compensated matching degree;

Combining the matching degree after the time delay compensation and the matching degree after the frequency domain compensation according to a preset proportion to generate comprehensive compensation matching degree of each candidate model;

And carrying out descending order sorting on the candidate model set according to the comprehensive compensation matching degree, selecting the highest-order candidate model as a final matching model, and generating an identification decision containing the final matching model.

In a second aspect, the present application provides a specific unmanned aerial vehicle model rapid identification system based on a radio frequency fingerprint library, comprising:

the receiving module is used for receiving radio frequency signals generated by the target unmanned aerial vehicle in flight, dynamically denoising the radio frequency signals and generating denoised radio frequency signals;

The matching module is used for separating combined data containing a signal intensity fluctuation mode, a phase jump frequency and a modulation distortion characteristic from the noise-reduced radio frequency signal, and carrying out layer-by-layer matching on the combined data and a pre-constructed radio frequency fingerprint library to generate a matching result;

The extraction module is used for extracting model-related parameters of the target unmanned aerial vehicle from the matching result, wherein the model-related parameters comprise signal propagation delay differences corresponding to specific obstacle reflection paths and frequency domain distortion association degrees corresponding to electromagnetic interference source distribution;

And the output module is used for generating an identification decision comprising the real-time environment adaptation factor based on the model association parameter and outputting a target unmanned model number result corresponding to the identification decision.

In a third aspect, an embodiment of the application provides a computing device, which comprises a processing component and a storage component, wherein the storage component stores one or more computer instructions, and the one or more computer instructions are used for being called and executed by the processing component to realize the specific unmanned aerial vehicle model rapid identification method based on the radio frequency fingerprint library.

In a fourth aspect, an embodiment of the present application provides a computer storage medium, where a computer program is stored, where the computer program is executed by a computer to implement a method for quickly identifying a specific unmanned aerial vehicle model based on a radio frequency fingerprint library according to the first aspect.

The embodiment of the application effectively solves the core problem of unmanned plane radio frequency signal characteristic distortion under complex electromagnetic interference and multi-obstacle reflection scenes through the collaborative design of dynamic noise reduction processing and layer-by-layer matching of combined data. The dynamic noise reduction processing is used for adaptively adjusting the filtering intensity based on real-time environment parameters (interference intensity and obstacle density), obviously inhibiting the influence of noise components on key characteristics such as a signal intensity fluctuation mode, a phase jump frequency and the like, generating combined data by separating and fusing multidimensional characteristics, and greatly improving the differentiation degree of fine signal differences by combining a layer-by-layer matching mechanism (progressive screening of intensity fluctuation, phase jump and modulation distortion) of a radio frequency fingerprint library, so that model identification with high precision and low time delay is realized in an open scene.

Further, the environment adaptation factor is decomposed into a time delay correction factor (associated obstacle reflection path) and a frequency domain correction factor (associated electromagnetic interference distribution), the time delay difference parameter and the frequency domain distortion parameter of the candidate model are respectively compensated and corrected twice, and the comprehensive matching degree is generated by combining the preset proportion, so that the recognition decision robustness under the environment fluctuation is ensured. The mechanism not only avoids the problem of insufficient real-time performance of traditional static threshold matching, but also realizes continuous optimization of scene self-adaption by incrementally updating the radio frequency fingerprint library, and finally achieves dual improvement of stability and accuracy of unmanned model identification in a complex open environment.

These and other aspects of the application will be more readily apparent from the following description of the embodiments.

Drawings

In order to more clearly illustrate the embodiments of the present application or the technical solutions of the prior art, the following description will briefly explain the drawings used in the embodiments or the description of the prior art, and it is obvious that the drawings in the following description are some embodiments of the present application, and other drawings can be obtained according to these drawings without inventive effort for a person skilled in the art.

Fig. 1 shows a flowchart of a specific unmanned aerial vehicle model quick identification method based on a radio frequency fingerprint library provided by the application;

fig. 2 shows a schematic structural diagram of a specific unmanned aerial vehicle model quick recognition system based on a radio frequency fingerprint library provided by the application;

FIG. 3 illustrates a schematic diagram of a computing device provided by the present application.

Detailed Description

In order to enable those skilled in the art to better understand the present application, the following description will make clear and complete descriptions of the technical solutions according to the embodiments of the present application with reference to the accompanying drawings.

In some of the flows described in the specification and claims of the present application and in the foregoing figures, a plurality of operations occurring in a particular order are included, but it should be understood that the operations may be performed out of order or performed in parallel, with the order of operations such as 101, 102, etc., being merely used to distinguish between the various operations, the order of the operations themselves not representing any order of execution. In addition, the flows may include more or fewer operations, and the operations may be performed sequentially or in parallel. It should be noted that, the descriptions of "first" and "second" herein are used to distinguish different messages, devices, modules, etc., and do not represent a sequence, and are not limited to the "first" and the "second" being different types.

In an open scene of complex electromagnetic interference and multi-obstacle reflection, the existing matching recognition scheme based on a static radio frequency fingerprint library faces serious challenges, wherein the static fingerprint library depends on reference data in a fixed experimental environment, the electromagnetic interference intensity and the obstacle reflection path which change in real time in a dynamic open scene are difficult to adapt, so that the offset of the signal characteristics of a target unmanned aerial vehicle and the reference fingerprint exceeds a preset threshold value to cause misjudgment, meanwhile, a fixed threshold matching mechanism cannot respond to dynamic fluctuation of the signal distortion degree, the threshold value is required to be repeatedly calibrated or the fingerprint library is updated, the real-time recognition is severely restricted, and in addition, the existing scheme lacks a cooperative mechanism of environment parameters (such as interference source distribution and obstacle position) and radio frequency fingerprint correction, so that recognition efficiency fluctuation in different scenes is obvious, and high-precision supervision requirements are difficult to meet.

Aiming at the problems, the application provides a specific unmanned aerial vehicle model quick identification method based on a radio frequency fingerprint library, which is characterized in that the inherent limitations of a static fingerprint library and a fixed threshold mechanism are broken through a closed loop technology link of dynamic noise reduction processing, combined data layered matching and environment adaptation factor dynamic compensation. The method comprises the steps of driving dynamic noise reduction processing through real-time environment parameters (interference intensity and obstacle density), adaptively restraining influence of noise components on core characteristics such as a signal intensity fluctuation mode and phase jump frequency, separating a noise-reduced radio frequency signal into multi-dimensional combined data, screening candidate models through layer-by-layer matching (intensity fluctuation, phase jump and modulation distortion) of a radio frequency fingerprint library, and carrying out dynamic compensation correction on a matching result by combining an environment adaptation factor. According to the scheme, through real-time cooperation of the environment parameters and the radio frequency fingerprint library (such as adjustment of delay difference weight based on the dynamic speed of the obstacle and correction of frequency domain distortion association degree based on the interference fluctuation amplitude), the signal characteristic offset is effectively eliminated, repeated calibration operation is avoided, identification instantaneity, accuracy and environmental robustness in a dynamic open scene are remarkably improved, and reliable technical support is provided for unmanned aerial vehicle supervision and countermeasures.

The following description of the embodiments of the present application will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present application, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the application without making any inventive effort, are intended to fall within the scope of the application.

Fig. 1 is a flowchart of a method for quickly identifying a specific unmanned aerial vehicle model based on a radio frequency fingerprint library according to an embodiment of the present application, as shown in fig. 1, the method includes:

step 101, receiving a radio frequency signal generated by a target unmanned aerial vehicle in flight, and performing dynamic noise reduction on the radio frequency signal to generate a noise-reduced radio frequency signal;

In the step, dynamic noise reduction processing refers to a process of adaptively adjusting a noise suppression strategy according to real-time environmental parameters (such as interference intensity distribution and obstacle density distribution), wherein the real-time interference intensity distribution represents electromagnetic interference energy distribution of different frequency bands in a monitoring area, the obstacle density distribution represents the density degree of obstacle reflection paths in a unit space, multistage series filtering refers to noise suppression operation of different frequency bands and amplitudes on radio frequency signals based on the environmental parameters, single-stage filtering refers to noise suppression of a single frequency band or amplitude, and signal integrity recovery processing refers to restoration of key signal characteristics which are possibly lost in the filtering process and related to flight states of unmanned aerial vehicles through a signal segment recombination and feature retention mechanism.

In this embodiment, first, the original radio frequency signals of the target unmanned aerial vehicle are synchronously captured by a multi-channel signal acquisition device deployed in the monitoring area, where the device covers different spatial positions to capture multi-path reflected signals. Then, the energy distribution (such as the interference intensity of 2.4GHz/5.8GHz frequency bands) of different frequency bands in the original signal is analyzed in real time, and the obstacle distribution density parameters are synchronously acquired through a laser radar or an infrared sensor. According to the combination relation of the interference intensity and the obstacle density (such as a high-interference and high-density scene), multi-stage series filtering is activated, wherein the first stage is used for filtering broadband noise outside a main frequency band, the second stage is used for carrying out amplitude threshold cutting on impulse noise, the third stage is used for carrying out frequency domain notch processing on residual noise, and if the interference intensity or the obstacle density is lower than the threshold value, only single-stage main frequency band-pass filtering is carried out. Finally, detecting the time domain waveform continuity of the filtered signal, identifying signal mutation or phase jump breaking points caused by filtering, performing interpolation restoration based on adjacent section signal characteristics (such as amplitude slope and phase continuity), reserving signal intensity mutation sections and phase jump sections related to the flight height and the attitude of the unmanned aerial vehicle, and generating a noise-reduced radio frequency signal.

For example, in an urban building group monitoring scene, a multichannel signal acquisition device deployed on the roof captures an original radio frequency signal of a target unmanned aerial vehicle, strong interference caused by a WiFi router is detected to exist in a 2.4GHz frequency band (the interference intensity distribution shows that an energy peak exceeds a threshold), and meanwhile, the laser radar feedback obstacle density distribution is that the reflection paths of high-rise buildings are dense (the density value is higher than the threshold). According to the method, the system activates multistage series filtering, namely, the first stage filters stray signals outside a 2.4GHz frequency band, the second stage cuts the amplitude of impulse noise caused by building reflection, and the third stage performs frequency domain notch on residual narrowband interference. The filtered signals are subjected to waveform continuity detection to find two phase jump breaking points, linear interpolation restoration is carried out based on the phase slopes of the front section and the rear section, and finally noise reduction signals retaining the relevant signal strength fluctuation of the fly height and the integral phase jump characteristics are generated for subsequent combined data separation and matching.

Step 102, separating combined data containing a signal intensity fluctuation mode, a phase jump frequency and a modulation distortion characteristic from a noise-reduced radio frequency signal, and carrying out layer-by-layer matching on the combined data and a pre-constructed radio frequency fingerprint library to generate a matching result;

In the step, the signal intensity fluctuation mode refers to a time sequence fluctuation rule of radio frequency signal intensity along with the change of the flight height and the gesture of the unmanned aerial vehicle, the phase jump frequency refers to the frequency and interval distribution characteristic of discontinuous jump of the signal phase in unit time, the modulation distortion characteristic refers to the distortion position and the form difference of a distortion interval of a radio frequency signal modulation waveform relative to a standard modulation mode, the combined data are multidimensional data sets formed by fusion of the three characteristics according to preset weights, the radio frequency fingerprint library stores reference characteristic data of unmanned aerial vehicles of different types in a typical scene, and the layer-by-layer matching refers to the sequential layered screening of candidate models through the signal intensity, the phase jump and the modulation distortion.

In this embodiment, the radio frequency signal after noise reduction is first divided into continuous signal segments according to a preset time window (e.g. 50 ms), so as to ensure that each signal segment contains a complete flight state period (e.g. climb, hover). For each signal segment, extracting a signal intensity fluctuation mode, namely aligning a signal intensity time sequence curve through a dynamic time warping algorithm, calculating fluctuation similarity between the signal intensity time sequence curve and a height change reference curve, extracting phase jump frequency, namely counting phase jump times in the signal segment, calculating a frequency distribution histogram of jump intervals through a sliding window, extracting a modulation distortion characteristic, namely comparing a signal segment modulation waveform with a standard waveform point by point, marking a distortion position and calculating a distortion interval length ratio. And then fusing the fluctuation mode, the frequency distribution and the distortion characteristics into combined data units according to preset weights (such as 5:3:2), and splicing the combined data units of all the signal segments into complete combined data. And then, inputting the combined data into a radio frequency fingerprint library, screening a candidate model set with the similarity of the signal strength fluctuation mode higher than a threshold (such as 80%) based on a similarity calculation method of dynamic time warping in the first layer matching, and generating a comprehensive matching degree of the candidate models through overlapping area calculation of phase jump frequency distribution and offset weighted summation of modulation distortion positions in the second layer matching, and generating a matching result according to descending order.

For example, in the urban building group monitoring scenario, the noise reduction radio frequency signal generated in step 101 is divided into 50ms time windows, and a signal segment corresponds to a flight phase of the unmanned aerial vehicle crossing the gap of the high-rise building. When the signal intensity fluctuation mode is extracted, the intensity curve of the signal intensity fluctuation mode shows periodic attenuation due to building shielding, after the intensity curve is aligned with a reference curve of a model number C in a fingerprint library in time sequence, the peak Gu Chazhi difference is smaller than 8%, the phase jump frequency statistics shows that the jump frequency of the section is 12 times/50 ms (the phase jump is increased due to multipath reflection), the overlapping area ratio of a frequency distribution histogram and the reference histogram of the model number C reaches 78%, the waveform distortion position is found to be concentrated on a modulation falling edge (caused by building reflection path delay) through modulation distortion feature analysis, and the distortion interval is 15%. After the combination data are fused according to preset weight (signal intensity fluctuation 5: phase jump 3: modulation distortion 2), the model C, D is screened out as a candidate set by the first layer matching, the weighting score of the phase jump overlapping area (78%) and the distortion offset (0.08) of the model C is 0.82, the model D is 0.65, the result of the model C being the highest matching degree is finally generated,

Step 103, extracting model-related parameters of the target unmanned aerial vehicle from the matching result, wherein the model-related parameters comprise signal propagation delay differences corresponding to specific obstacle reflection paths and frequency domain distortion association degrees corresponding to electromagnetic interference source distribution;

In the step, the signal propagation delay difference refers to a deviation value between the signal propagation time delay of the target unmanned aerial vehicle in the current obstacle distribution scene and the reference time delay of the reference model in the radio frequency fingerprint library in the same reflection path, and the frequency domain distortion association degree refers to the coincidence degree between the signal frequency domain distortion position of the target unmanned aerial vehicle in the electromagnetic interference source distribution scene and the reference distortion position of the reference model in the same interference condition.

In this embodiment, based on the matching result (the candidate model set and the matching degree ranking) generated in step 102, the radio frequency fingerprint reference data corresponding to the highest candidate model is extracted. Firstly, acquiring the propagation delay of a reference signal of the model under a typical obstacle reflection path from radio frequency fingerprint reference data (for example, the reference delay of a high-rise building reflection path is 1.2 ms), and calculating the difference value (0.3 ms) of the propagation delay and the actual delay (for example, 1.5 ms) of the same reflection path of the target unmanned aerial vehicle under the current scene by comparing the actual delay of the target unmanned aerial vehicle as a signal propagation delay difference parameter. And secondly, acquiring a reference frequency domain distortion position (such as a 2.4GHz frequency band distortion interval caused by WiFi interference) of the model under typical electromagnetic interference distribution from radio frequency fingerprint reference data, and calculating the superposition ratio (such as 85%) of the actual distortion position (such as 2.4-2.42GHz interval distortion) of the target unmanned aerial vehicle under the current interference source distribution and the reference distortion interval as a frequency domain distortion association degree parameter by analyzing the actual distortion position (such as the distortion of the target unmanned aerial vehicle. And finally, integrating the time delay difference parameter and the frequency domain association degree parameter into model association parameters for the calculation of subsequent environment adaptation factors.

For example, in the city building group monitoring scenario, the matching result of step 102 determines model C as the highest-matching candidate model. And extracting the reference time delay (1.2 ms) of the model C under the high-density building reflection path from the radio frequency fingerprint reference data, comparing the actual time delay (the actual time delay is 1.55ms due to multipath reflection enhancement) of the same path in the current scene, and calculating to obtain the signal propagation time delay difference parameter of 0.35ms. Meanwhile, a standard distortion interval (2.4-2.41 GHz) of the model C under WiFi interference of a 2.4GHz frequency band is extracted from the reference data, an actual distortion interval of a target unmanned aerial vehicle in a current scene is 2.4-2.415GHz (the interval is expanded by 5MHz due to the distribution change of an interference source), and the superposition ratio is calculated to be 80% (the superposition interval is 10 MHz/the total distortion interval is 15 MHz) and is used as a frequency domain distortion association parameter. After the parameters are integrated, the model association parameters comprise delay difference of 0.35ms and frequency domain association degree of 80 percent,

104, Generating an identification decision containing a real-time environment adaptation factor based on the model-related parameters, and outputting a target unmanned model number result corresponding to the identification decision;

In the step, the real-time environment adaptation factor refers to a weight correction coefficient generated based on current scene dynamic parameters (the change rate of obstacle position and the fluctuation amplitude of interference source intensity) and used for compensating the environment offset of the matching result of the radio frequency fingerprint library, the dynamic compensation mechanism refers to a process of secondarily adjusting the matching degree of the candidate model according to the environment adaptation factor, and the consistency verification mechanism refers to a step of carrying out logic consistency verification on the identification of the final matching model and the pre-stored model characteristics in the radio frequency fingerprint library.

In this embodiment, the signal propagation delay difference and the frequency domain distortion correlation degree are extracted from the model correlation parameters, and input to the environment adaptation factor calculation model. The model generates a time delay correction factor and a frequency domain correction factor based on real-time tracking data (the change rate of the obstacle position is 5 m/s) and fluctuation data through a preset rate-weight mapping table and an amplitude-weight piecewise function. And then, dynamically compensating the matching degree of the correction factor and the candidate model, namely multiplying the time delay correction factor by the initial matching degree and the time delay difference parameter of the model C to obtain the time delay compensation matching degree, and multiplying the frequency domain correction factor by the frequency domain association degree parameter to obtain the frequency domain compensation matching degree. And combining the two compensation results according to a preset proportion to generate the comprehensive compensation matching degree. And sorting according to the comprehensive matching degree, wherein the model C is still the highest matching result. And finally, carrying out consistency verification on the identification (such as a serial number UA-C-001) of the model C and model characteristics (a flying height-signal intensity curve and a standard modulation waveform) prestored in a radio frequency fingerprint library, confirming that the signal intensity fluctuation mode error is <5% and the modulation distortion coincidence degree is more than 75%, and outputting a model C as a target unmanned model number result after verification is passed.

For example, in the urban building group monitoring scenario, the model-related parameters (delay difference 0.35ms, frequency domain correlation 80%) extracted in step 103 are input into the environment adaptation factor calculation model. According to the real-time barrier position change rate of 5m/s (exceeding a threshold value of 4 m/s), triggering time delay weight linear lifting coefficient of 1.2, superposition basic weight of 1.0 to generate time delay correction factor of 1.2, meanwhile, electromagnetic interference source fluctuation amplitude of + -3 dB (exceeding a threshold value of 2 dB), triggering frequency domain weight according to a difference of 1dB to gradually increase gradient of 0.1/dB, and generating frequency domain correction factor of 1.1. The initial matching degree of model C is calculated by compensating 0.82, namely time delay compensation matching degree=0.82×1.2× (1-0.35 ms/reference time delay 1.2 ms) =0.82×1.2×0.708≡0.698, and frequency domain compensation matching degree=0.698×1.1× (80%/100%) =0.698×1.1×0.8≡0.614. The combined comprehensive matching degree is still higher than that of other candidate models by 0.614. And checking the mark of the model C with pre-stored characteristics in a fingerprint library, namely, 3 percent of measured signal intensity fluctuation mode error (< 5 percent) and 78 percent of modulation distortion coincidence degree (> 75 percent), and outputting the model C as a final recognition result after the verification is passed, wherein the model C is consistent with the actual model of the unmanned aerial vehicle in the scene.

In order to solve the problem of interference of noise components of the radio frequency signal caused by complex electromagnetic interference and obstacle reflection, and improve the noise reduction quality of the signal in a dynamic open scene, in some embodiments, according to step 102, a radio frequency signal generated by a target unmanned aerial vehicle in flight is received, and dynamic noise reduction processing is performed on the radio frequency signal, so as to generate a noise-reduced radio frequency signal, including:

step 201, capturing an original radio frequency signal of a target unmanned aerial vehicle in real time through a multichannel signal acquisition device deployed in a monitoring area, wherein the original radio frequency signal comprises a noise component caused by electromagnetic interference sources and obstacle reflections in a current scene;

In the step, the multichannel signal acquisition equipment refers to a plurality of radio frequency receiving modules distributed in a monitoring area and used for synchronously acquiring radio frequency signals emitted by the unmanned aerial vehicle from different spatial positions, and noise components in the original radio frequency signals comprise broadband noise caused by electromagnetic interference sources (such as WiFi routers and base stations) and impulse noise formed by multipath interference caused by obstacle reflection.

In this embodiment, in the urban building group monitoring scenario, the multi-channel signal acquisition devices (covering the east, west and south directions) deployed on three adjacent roofs are simultaneously started, and the radio frequency signals emitted by the target unmanned aerial vehicle in the flight process are captured in real time. The east side equipment receives strong broadband noise with 2.4GHz frequency band in the signal due to the fact that the east side equipment is close to a WiFi hot spot of a commercial area, the west side equipment receives impulse noise (signal burrs with abrupt amplitude change) of multipath propagation due to reflection of a high-layer glass curtain wall, and the south side equipment receives weak noise components in the signal due to less shielding. After the signals of all channels are calibrated by the synchronous clock, the signals are summarized to the central processing unit to form an original radio frequency signal data set containing space difference characteristics.

Step 202, monitoring environmental parameters in the original radio frequency signal, wherein the environmental parameters comprise real-time interference intensity distribution and obstacle density distribution;

in the step, the real-time interference intensity distribution refers to the spatial distribution characteristics of electromagnetic interference energy in different frequency bands, the spatial distribution characteristics are extracted through spectrum analysis, and the obstacle density distribution refers to the number and the spatial density of obstacle reflection paths in a unit area, and the number and the spatial density are measured through a laser radar or an infrared sensor.

In the embodiment, the original radio frequency signals summarized in the step 201 are subjected to multichannel spectrum analysis, an east channel signal is subjected to fast Fourier transform to generate an energy distribution diagram of a frequency band of 2.4-2.48GHz, a WiFi interference intensity peak value is concentrated at 2.412GHz (-22 dBm), a west channel signal is subjected to time domain amplitude statistics to identify a dense interval (20-30 mutation per second) of impulse noise, and meanwhile, a laser radar scans a west glass curtain wall area of a building group, and the feedback obstacle reflection path density is 5 pieces per square meter. The data are integrated to generate environmental parameters, wherein the real-time interference intensity distribution is marked as {2.4GHz band: strong, 5.8GHz band: weak }, and the obstacle density distribution is marked as { west area: high, east/south area: medium }.

Step 203, selecting a signal filtering intensity level according to the environmental parameter, if the real-time interference intensity is higher than a set range and the obstacle density is higher than a set threshold, performing layer-by-layer noise suppression on the original radio frequency signal by adopting multistage series filtering, and if the real-time interference intensity is lower than the set range, performing noise suppression on the original radio frequency signal by adopting single-stage filtering to obtain a filtered radio frequency signal;

In this step, multistage series filtering refers to the operations of frequency band isolation, amplitude truncation and frequency domain notch being sequentially performed based on different noise types (wideband noise, impulse noise, etc.), single stage filtering refers to bandpass or notch processing of a single noise type (such as main frequency band interference), and signal filtering intensity level refers to a strategy of dynamically selecting a filtering level and a suppression amplitude according to environmental parameters.

In this embodiment, according to the environmental parameters monitored in step 202 (the real-time interference intensity distribution is marked as 2.4GHz band strong interference, the obstacle density distribution is marked as west high density), it is determined that the multistage series filtering condition (interference intensity > threshold, obstacle density > threshold) is satisfied. Firstly, performing first-stage frequency band isolation filtering on an original radio frequency signal of an east side channel, namely cutting off spurious signals (keeping 2.4-2.48 GHz) outside a 2.4GHz main frequency band and restraining broadband noise caused by WiFi, secondly, performing second-stage amplitude cut-off filtering on a west side channel signal, namely detecting an amplitude mutation interval (> 30 dB) of impulse noise, replacing an adjacent section mean value to restrain multipath reflection noise, and finally performing third-stage frequency domain notch filtering on residual narrowband interference, namely generating a notch window at WiFi hot spot frequency points such as 2.412GHz and 2.422GHz, and further filtering residual interference. The south side channel performs only a single stage of main band pass filtering due to low interference strength (< threshold).

Step 204, performing signal integrity recovery processing on the filtered radio frequency signal to generate a noise-reduced radio frequency signal;

In the step, the signal integrity recovery processing refers to repairing signal waveform distortion caused by filtering operation, including phase jump fracture repair, amplitude jump loss compensation and waveform continuity reconstruction, and the key feature retention refers to recovering signal strength jump segments and phase jump segments which are strongly related to the flight state (such as altitude and attitude) of the unmanned aerial vehicle.

In the embodiment, the waveform integrity of the multichannel signal after filtering in step 203 is detected, namely, the eastern side channel filtering signal breaks at the boundary of 2.48GHz due to frequency band truncation, a linear interpolation signal is generated through adjacent frequency band phase slope calculation to repair the jump, the western side channel filtering signal loses part of amplitude mutation due to amplitude truncation, a mutation peak-valley value is reconstructed based on the amplitude change trend of the front section and the rear section, the southern side channel signal has good integrity, and only the waveform of a transition area needs to be smoothed. After the processing is completed, the intensity abrupt segment (for example, the intensity of the hovering stage is reduced by 20 dB) and the complete phase jump segment (for example, the jump interval of the climbing stage is shortened to 5 ms) which are related to the flying height in each channel signal are extracted, and the noise-reduced radio frequency signals are generated by combining.

In order to improve matching robustness of the multi-dimensional signal features in a complex scene, in some embodiments, according to the step 102, combined data including a signal strength fluctuation mode, a phase jump frequency and a modulation distortion feature is separated from the radio frequency signal after noise reduction, and the combined data is matched with a pre-constructed radio frequency fingerprint library layer by layer to generate a matching result, which includes:

Step 301, dividing the noise-reduced radio frequency signal into a plurality of signal segments according to a preset time window, and extracting the fluctuation mode of the signal intensity along with the change of the flying height, the frequency distribution of the phase jump interval and the position characteristics of modulation waveform distortion from each signal segment;

In the step, the time window division refers to dividing a continuous radio frequency signal into signal segments with fixed duration for analyzing the periodic characteristics of a flight state, the signal strength fluctuation mode refers to the time sequence peak-to-valley change rule generated by the rise and fall of the signal strength along with the flight height of the unmanned aerial vehicle, the phase jump frequency distribution refers to the frequency of discontinuous jump in the signal segments and the time interval statistical characteristics thereof, and the modulation waveform distortion position characteristics refer to the position and the length ratio of a distortion section of a signal modulation waveform relative to a standard waveform on a time axis.

In the embodiment, the radio frequency signal after noise reduction is firstly divided into continuous signal segments of a preset time window, each signal segment corresponds to different stages of the unmanned aerial vehicle flight state, a time sequence alignment algorithm is used for matching a signal intensity curve with a flight height change curve, similarity of peak-to-valley positions, amplitude differences and change rates is analyzed to extract fluctuation mode features, then phase jump times in each signal segment are counted, distribution characteristics of jump intervals are calculated through a sliding window to generate frequency distribution features, meanwhile, signal segment modulation waveforms are compared with standard waveforms point by point, the positions and length proportion of distortion segments are marked to extract position features, and finally each signal segment outputs feature vectors containing fluctuation modes, frequency distribution and distortion positions.

Step 302, fusing the fluctuation mode, the frequency distribution and the position characteristics into combined data according to a preset weight proportion;

In the step, the preset weight proportion refers to the step of giving different importance weights (such as signal strength fluctuation weight 5, phase jump frequency weight 3 and modulation distortion weight 2) to three types of features according to scene requirements, wherein the combined data is a multidimensional data set obtained by weighting and fusing three types of feature vectors and is used for matching calculation of a radio frequency fingerprint library.

In this embodiment, the extracted fluctuation mode, frequency distribution and position feature are respectively normalized and weighted and fused according to a preset weight proportion, wherein the fluctuation mode calculates the matching degree with a reference curve through similarity difference, the frequency distribution determines the matching degree with a reference distribution through overlapping area analysis, the position feature evaluates the consistency of a distortion section through deviation tolerance, a combined data unit is generated after weighted summation of the three, and the combined data units of all signal sections are spliced into complete combined data according to time sequence and input into a radio frequency fingerprint library for matching.

Step 303, carrying out similarity calculation on the combined data and a first layer signal intensity fluctuation mode reference set stored in a radio frequency fingerprint library in sequence, and screening a candidate model set with similarity higher than a first threshold;

In the step, a first layer of signal intensity fluctuation mode reference set refers to a reference fluctuation curve of signal intensity of each type unmanned aerial vehicle under a typical scene, which is prestored in a radio frequency fingerprint library and changes along with flying height, and similarity calculation refers to quantification of matching degree of a target signal fluctuation mode and the reference curve through a dynamic time warping or waveform difference accumulation algorithm.

In this embodiment, the wave pattern feature vectors of all signal segments in the combined data generated in step 302 are sequentially aligned with the first layer reference wave curves of all models in the radio frequency fingerprint library in time sequence and calculated in a difference accumulation manner to generate a similarity score of each model, models with similarity scores higher than a preset threshold are screened to form candidate sets, and the corresponding second layer phase jump frequency reference set and third layer modulation distortion reference set are loaded to the next matching stage.

Step 304, performing joint matching on the frequency distribution and the position feature in the combined data and a second layer phase jump frequency reference set and a modulation distortion reference set corresponding to the candidate model set to generate a matching result containing the matching degree sequence of the candidate models;

In the step, the second layer of phase jump frequency reference set refers to a phase jump interval distribution histogram of a candidate model under a typical scene, the modulation distortion reference set refers to reference data of modulation waveform distortion positions and length duty ratios of the candidate model under the same interference condition, and the joint matching refers to weighted comprehensive evaluation of frequency distribution overlapping area calculation and distortion position offset analysis.

In this embodiment, second-layer reference data (phase jump frequency distribution histogram) and third-layer reference data (modulation distortion position) of each model are extracted from a candidate model set, overlapping area calculation is performed on frequency distribution features and the phase jump reference histogram in target combination data, offset analysis is performed on position features and modulation distortion references, weighted summation is performed on two results according to preset weights, comprehensive matching degree of each candidate model is generated, and final matching results are formed according to descending order.

In order to solve the differential effect of the obstacle reflection path and the electromagnetic interference source distribution on the signal characteristics in the dynamic scene, in some embodiments, according to step 103, a model association parameter of the target unmanned aerial vehicle is extracted from the matching result, where the model association parameter includes a signal propagation delay difference corresponding to a specific obstacle reflection path and a frequency domain distortion association degree corresponding to the electromagnetic interference source distribution, and the method includes:

step 401, according to the matching degree sequence of the candidate models in the matching result, extracting the radio frequency fingerprint reference data corresponding to the highest candidate model;

In the step, the candidate model matching degree sorting refers to a model list which is generated through a hierarchical matching process (signal strength fluctuation, phase jump and modulation distortion) and is arranged in descending order according to the comprehensive matching degree, and the radio frequency fingerprint reference data refers to reference characteristic data of a target model pre-stored in a radio frequency fingerprint library under a typical scene, wherein the reference characteristic data comprises signal propagation delay, phase jump frequency and modulation distortion position parameters.

In this embodiment, based on the matching result (such as model C > model E > model F) generated in step 304, the radio frequency fingerprint reference data corresponding to the highest candidate model (model C) is extracted, including the propagation delay of the reference signal, the distribution histogram of the phase jump interval and the reference interval of the modulation waveform distortion position of the model in the typical obstacle distribution scene, so as to provide a reference comparison basis for extracting the related parameters of the subsequent models.

Step 402, acquiring signal propagation delay difference parameters associated with obstacle distribution in a current scene from the radio frequency fingerprint reference data;

In the step, the signal propagation delay difference parameter refers to a deviation value between an actual signal propagation delay of the target unmanned aerial vehicle under the current obstacle distribution and a reference delay in the same scene in the radio frequency fingerprint reference data, and is used for quantifying the influence degree of the environmental dynamic change on the signal propagation path.

In this embodiment, the propagation delay (such as reference delay) of the reference signal in the typical dense high-rise building scene is obtained from the reference data of the radio frequency fingerprint of the model C, and the difference between the two is calculated as a delay difference parameter by comparing the actual delay (such as actual delay) under the same obstacle distribution condition in the current scene, so as to reflect the influence of the dynamic change (such as newly added temporary obstacle) of the current obstacle position on the propagation path of the signal.

Step 403, obtaining a frequency domain distortion association parameter associated with electromagnetic interference source distribution from the radio frequency fingerprint reference data;

In the step, the frequency domain distortion association parameter refers to the superposition ratio between the actual signal frequency domain distortion position of the target unmanned aerial vehicle under the current electromagnetic interference source distribution and the reference distortion position under the same interference condition in the radio frequency fingerprint reference data, and is used for quantifying the interference degree of the electromagnetic interference dynamic change on the signal frequency domain characteristics.

In this embodiment, the reference frequency domain distortion position (for example, a specific frequency band interval) of the model C under a typical electromagnetic interference scene (for example, wiFi hot spot dense distribution) is extracted from the radio frequency fingerprint reference data, and the superposition ratio of the actual distortion interval and the reference interval is calculated by comparing the actual distortion position (for example, frequency band expansion or offset) caused by the electromagnetic interference source distribution (for example, a newly added temporary communication base station) in the current scene, so as to generate a frequency domain distortion association parameter, and reflect the offset influence of the current interference fluctuation on the signal characteristics.

Step 404, integrating the signal propagation delay difference parameter and the frequency domain distortion association parameter into model association parameters of the target unmanned aerial vehicle;

In the step, the model association parameter refers to a composite parameter fusing signal propagation delay difference (reflecting obstacle dynamic influence) and frequency domain distortion association degree (reflecting electromagnetic interference dynamic influence), and is used for quantifying the joint effect of current scene environment fluctuation on the signal characteristics of the target unmanned aerial vehicle.

In this embodiment, the signal propagation delay difference parameter (such as delay deviation caused by new addition of an obstacle) extracted in step 402 and the frequency domain distortion association parameter (such as the reduction of the overlap ratio caused by interference expansion) generated in step 403 are integrated according to a preset format (such as a vector or a matrix), so as to form a composite parameter set including delay difference weights and frequency domain association weights, input is provided for dynamic calculation of a subsequent real-time environment adaptation factor, and reference data correction logic of a target model in a radio frequency fingerprint library is associated.

In order to improve the real-time performance and accuracy of the recognition decision under dynamic environment fluctuation, in some embodiments, according to the step 104, generating a recognition decision including a real-time environment adaptation factor based on the model-related parameter, and outputting a target unmanned aerial vehicle model result corresponding to the recognition decision, including:

Step 501, acquiring real-time tracking data of dynamic change of the position of an obstacle in a current scene and real-time fluctuation data of the strength of an electromagnetic interference source;

In the step, real-time tracking data refer to the moving speed and distribution density change information of the position of the obstacle, which are acquired in real time through a multi-source sensor (such as a laser radar and a camera), and real-time fluctuation data refer to the dynamic change characteristics of the energy intensity of an electromagnetic interference source, which is monitored through spectrum analysis equipment, in the frequency domain and the time domain.

In the embodiment, the laser radar array deployed in the monitoring area scans the position of the obstacle in real time to generate a thermal distribution map comprising the moving speed (such as the displacement of the obstacle in unit time) and the density change, meanwhile, the distributed spectrum monitoring network is used for sampling the frequency band energy of an electromagnetic interference source (such as a WiFi router and a base station) to generate the fluctuation amplitude (such as the peak energy change range) and the time-sequence fluctuation trend of the interference intensity of each frequency point, and the data are aligned in time and space to form an environment parameter set which is strongly related to the dynamic state of the current scene to provide input for the adjustment of the weight coefficient.

Step 502, adjusting a weight coefficient of a signal propagation delay difference parameter according to the real-time tracking data, and adjusting a weight coefficient of a frequency domain distortion association parameter according to the real-time fluctuation data;

In this step, the weight coefficient adjustment refers to adaptively correcting the decision weight of the model-related parameters (delay difference, frequency domain correlation) based on the environmental dynamic parameters (obstacle movement rate, interference fluctuation amplitude) to enhance the response sensitivity to scene changes.

In the embodiment, according to the comparison relation between the moving speed of the obstacle in the real-time tracking data and the preset speed threshold, a linear enhancement rule is adopted to adjust the weight coefficient of the time delay difference parameter, if the speed exceeds the threshold, the weight is increased according to the ratio of the speed to the threshold, meanwhile, according to the difference value between the fluctuation amplitude of the interference intensity in the real-time fluctuation data and the preset amplitude threshold, a piecewise increment rule is adopted to adjust the weight coefficient of the frequency domain association parameter, if the amplitude exceeds the threshold, the weight coefficient gradient is increased according to the difference value piecewise, and finally the updated weight of the environment adaptation is generated.

Step 503, inputting the adjusted signal propagation delay difference parameter and the adjusted frequency domain distortion association parameter to an environment adaptation factor calculation model, and generating a real-time environment adaptation factor for correcting the matching degree of the candidate model in the radio frequency fingerprint library;

In the step, the environment adaptation factor calculation model refers to calculation logic for generating a dynamic correction factor by carrying out weighted fusion on signal propagation delay difference parameters and frequency domain distortion association parameters based on dynamic weight coefficients (delay difference weights and frequency domain association weights), wherein the real-time environment adaptation factor is a correction coefficient for quantifying the influence of current environment fluctuation on a matching result.

In this embodiment, the time delay difference weight coefficient and the frequency domain association weight coefficient adjusted in step 502 are respectively weighted and overlapped with corresponding time delay difference parameters (such as time delay deviation caused by movement of an obstacle) and frequency domain distortion association parameters (such as reduction of overlap ratio caused by interference expansion), so as to generate a time delay correction factor and a frequency domain correction factor, and the time delay correction factor and the frequency domain correction factor are combined into a real-time environment adaptation factor through a preset fusion rule (such as 60% of the time delay correction factor and 40% of the frequency domain correction factor), so as to dynamically correct the matching degree of candidate models in the radio frequency fingerprint library.

Step 504, dynamically compensating the matching degree of the candidate model in the matching result according to the real-time environment adaptation factor, and generating an identification decision containing the final matching model;

In the step, dynamic compensation refers to performing association calculation on the real-time environment adaptation factor and the initial matching degree of the candidate model to eliminate matching degree deviation caused by environment fluctuation, and the recognition decision is a final model judgment result generated by comprehensive matching degree sequencing after compensation.

In this embodiment, the initial matching degree (for example, the initial matching degree of model C is 0.82) of the candidate models is extracted from the matching result, the real-time environment adaptation factor, the initial matching degree, the time-domain correlation degree parameter and the frequency-domain correlation degree parameter are subjected to step-by-step product compensation to generate the time-domain compensation matching degree and the frequency-domain compensation matching degree, the two are combined according to a preset proportion to form the comprehensive compensation matching degree, the candidate models are sorted in descending order, the model with the highest matching degree is selected as the final result, and consistency verification is performed with model features in the radio frequency fingerprint library to ensure output reliability.

Step 505, performing consistency verification on the model identification of the final matching model in the identification decision and the model feature description prestored in the radio frequency fingerprint library, and outputting the final matching model as a target unmanned aerial vehicle model number result when the consistency verification passes;

In the step, the consistency verification refers to the process of carrying out multidimensional comparison on the actually measured characteristic data (such as a signal intensity fluctuation mode and a modulation distortion position) of the finally matched model and the model reference characteristic pre-stored in the radio frequency fingerprint library to verify the logic consistency of the model reference characteristic data, and the model characteristic description refers to the unique identification of the unmanned aerial vehicle model and the corresponding reference characteristic parameter set (such as a standard fluctuation curve and a modulation waveform template) stored in the radio frequency fingerprint library.

In the embodiment, pre-stored model feature description of a final matching model is extracted from a radio frequency fingerprint library, the pre-stored model feature description comprises a standard signal intensity fluctuation curve, a phase jump interval distribution histogram and a modulation waveform template, an actually measured signal intensity fluctuation mode of the model C is aligned with the standard fluctuation curve in time sequence, peak-valley position errors are calculated, meanwhile interval coincidence analysis is carried out on actually measured modulation distortion positions and the template, if the fluctuation mode errors are smaller than a preset threshold value and the modulation distortion coincidence degree is higher than the minimum requirement, consistency verification is judged to pass, a model result is output, otherwise abnormal alarm is triggered, and a matching process is executed again.

In order to solve the differential requirements of the dynamic change of the obstacle position and the fluctuation of the interference source intensity on the weight coefficient, in some embodiments, according to the step 502, the adjusting the weight coefficient of the signal propagation delay differential parameter according to the real-time tracking data, and according to the real-time fluctuation data, the adjusting the weight coefficient of the frequency domain distortion association parameter includes:

step 601, dynamically adjusting a weight coefficient of a signal propagation delay difference parameter according to the dynamic rate of the position change of the obstacle in the real-time tracking data, and linearly enhancing the weight coefficient according to the ratio of the dynamic rate to a preset rate threshold when the dynamic rate exceeds the preset rate threshold to generate an adjusted delay weight coefficient;

in this embodiment, the dynamic rate (for example, the displacement of the obstacle in a unit time) of the position change of the obstacle is extracted by tracking the data in real time, and is compared with a preset rate threshold, if the dynamic rate exceeds the threshold, the weight coefficient of the time difference parameter is linearly enhanced according to the ratio of the rate to the threshold, for example, when the rate exceeds the threshold, the weight coefficient is positively adjusted according to the rate exceeding ratio, and if the rate does not exceed the threshold, the original weight coefficient is kept unchanged, so that the actual influence of the movement of the obstacle on the signal propagation path is ensured to be reflected by the weight dynamic of the time difference parameter.

Step 602, dynamically adjusting a weight coefficient of a frequency domain distortion association parameter according to a fluctuation amplitude of electromagnetic interference source intensity in the real-time fluctuation data, and when the fluctuation amplitude exceeds a preset amplitude threshold, increasing the weight coefficient in a segmented manner according to a difference value between the fluctuation amplitude and the amplitude threshold to generate an adjusted frequency domain weight coefficient;

In this embodiment, the fluctuation amplitude (such as the interference energy peak Gu Chazhi) of the electromagnetic interference source intensity is extracted through the real-time fluctuation data, and compared with a preset amplitude threshold, if the fluctuation amplitude exceeds the threshold, the weighting coefficient of the frequency domain distortion association degree parameter is increased in a segmented manner according to the difference value between the amplitude and the threshold, for example, the weighting coefficient is increased in a gradient manner every time a certain amplitude interval exceeds the threshold, and if the fluctuation amplitude does not exceed the threshold, the original weighting coefficient is maintained, so that the self-adaptive adjustment of the frequency domain parameter weight along with the interference fluctuation degree is ensured.

To solve the limitation of the environmental adaptation factor in correcting the matching result in a single dimension, in some embodiments, according to the step 504, the dynamic compensation is performed on the matching degree of the candidate model in the matching result according to the real-time environmental adaptation factor, and the generating the recognition decision including the final matching model includes:

step 701, decomposing the real-time environment adaptation factor into a time delay correction factor and a frequency domain correction factor;

In the embodiment, the real-time environment adaptation factor is decomposed into a time delay correction factor and a frequency domain correction factor through a preset weight separation rule, wherein the time delay correction factor is generated by a dynamic weight adjustment result of a signal propagation time delay difference parameter and reflects the influence of the dynamic change of the obstacle position on the matching degree, the frequency domain correction factor is generated by a dynamic weight adjustment result of a frequency domain distortion association degree parameter and reflects the correction requirement of electromagnetic interference fluctuation on the matching degree, and the time delay correction factor and the frequency domain correction factor are decoupled from the environment adaptation factor through a weighted proportion distribution mechanism to form independent correction parameters so as to support step-by-step compensation logic.

Step 702, extracting an initial matching degree of a candidate model set from the matching result, and multiplying the delay correction factor by a signal propagation delay difference parameter corresponding to each candidate model in the initial matching degree item by item to generate a matching degree after delay compensation;

In this embodiment, the initial matching degree of the candidate model is extracted from the matching result (based on the hierarchical matching score of the signal strength fluctuation mode and the phase jump frequency), the time delay correction factor and the signal propagation time delay difference parameter (such as the time delay deviation value caused by the obstacle dynamics) corresponding to each candidate model are subjected to item-by-item product operation, and the matching degree after the time delay compensation is generated, wherein the larger the time delay difference parameter (the obvious influence of the obstacle movement) is, the stronger the compensation effect of the time delay correction factor on the matching degree is, so that the matching priority of the environment dynamic sensitive model is improved.

Step 703, multiplying the frequency domain correction factor by a frequency domain distortion association parameter corresponding to each candidate model in the time delay compensated matching degree, so as to generate a frequency domain compensated matching degree;

in this embodiment, frequency domain distortion association parameters (reflecting the coincidence degree of signal distortion and a reference under electromagnetic interference fluctuation) of each model are extracted from a candidate model set, a frequency domain correction factor (weight adjustment result triggered by interference fluctuation amplitude) and the frequency domain association parameters of each model are subjected to product-by-product operation to generate a matching degree after frequency domain compensation, wherein the lower the frequency domain association parameter (distortion coincidence degree is reduced due to interference), the stronger the compensation effect of the frequency domain correction factor on the matching degree is, so that the matching weight of an anti-interference model is preferentially improved.

Step 704, combining the time delay compensated matching degree and the frequency domain compensated matching degree according to a preset proportion to generate comprehensive compensation matching degree of each candidate model;

In this embodiment, a combination weight ratio of time delay compensation and frequency domain compensation is preset according to scene dynamic characteristics (for example, the time delay occupies 60% and the frequency domain occupies 40%), the matching degree after time delay compensation (reflecting the influence of obstacle movement) and the matching degree after frequency domain compensation (reflecting the influence of interference fluctuation) are weighted and combined to generate a comprehensive compensation matching degree, the combined correction capability of complex environment dynamics is enhanced through a weighted combination mechanism, and finally the candidate models are ordered in descending order according to the comprehensive matching degree, and an optimal recognition result is output.

Step 705, sorting the candidate model sets in descending order according to the comprehensive compensation matching degree, selecting the highest-sorting candidate model as the final matching model, and generating an identification decision containing the final matching model;

in the step, the comprehensive compensation matching degree refers to the candidate model comprehensive score generated by combining the preset proportion after the step-by-step compensation of the time delay correction factor and the frequency domain correction factor, the descending order refers to the candidate models which are arranged from high to low according to the comprehensive compensation matching degree, and the identification decision is a decision result data set containing the final matching model identification and the matching degree verification information.

In the embodiment, the system carries out numerical comparison on the comprehensive compensation matching degree of each model in the candidate model set, generates a sorting list according to a descending rule, selects the model with the highest ranking as the final matching model, and integrates the identification of the final matching model, the comprehensive compensation matching degree, the time delay/frequency domain compensation parameter and the environment adaptation factor into structural identification decision data for subsequent consistency verification and result output, thereby ensuring the traceability and reliability of the decision process.

Fig. 2 is a schematic structural diagram of a specific unmanned aerial vehicle model quick recognition system based on a radio frequency fingerprint library according to an embodiment of the present application, where, as shown in fig. 2, the system includes:

The specific unmanned aerial vehicle model quick recognition system based on the radio frequency fingerprint library shown in fig. 2 may execute the specific unmanned aerial vehicle model quick recognition method based on the radio frequency fingerprint library shown in the embodiment shown in fig. 1, and its implementation principle and technical effects are not repeated. The specific manner in which the modules and units perform the operations in the specific unmanned aerial vehicle model quick recognition system based on the radio frequency fingerprint library in the above embodiments has been described in detail in the embodiments related to the method, and will not be described in detail herein.

In one possible design, a specific unmanned aerial vehicle model quick identification system based on a radio frequency fingerprint library of the embodiment of fig. 2 may be implemented as a computing device, which may include a storage component 31 and a processing component 32, as shown in fig. 3;

the storage component 31 stores one or more computer instructions for execution by the processing component 32.

The processing component 32 is configured to perform a method for quickly identifying a specific model of a drone based on a library of rf fingerprints according to the embodiment of fig. 1.

Wherein the processing component 32 may include one or more processors to execute computer instructions to perform all or part of the steps of the methods described above. Of course, the processing component may also be implemented as one or more Application Specific Integrated Circuits (ASICs), digital Signal Processors (DSPs), digital Signal Processing Devices (DSPDs), programmable Logic Devices (PLDs), field Programmable Gate Arrays (FPGAs), controllers, microcontrollers, microprocessors or other electronic elements for executing the methods described above.

The storage component 31 is configured to store various types of data to support operations at the terminal. The memory component may be implemented by any type or combination of volatile or nonvolatile memory devices such as Static Random Access Memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic or optical disk.

Of course, the computing device may necessarily include other components as well, such as input/output interfaces, display components, communication components, and the like.

The input/output interface provides an interface between the processing component and a peripheral interface module, which may be an output device, an input device, etc.

The communication component is configured to facilitate wired or wireless communication between the computing device and other devices, and the like.

The computing device may be a physical device or an elastic computing host provided by the cloud computing platform, and at this time, the computing device may be a cloud server, and the processing component, the storage component, and the like may be a base server resource rented or purchased from the cloud computing platform.

The embodiment of the application also provides a computer storage medium which stores a computer program, and the computer program can realize the specific unmanned aerial vehicle model rapid identification method based on the radio frequency fingerprint library in the embodiment shown in the figure 1 when being executed by a computer.

It will be clear to those skilled in the art that, for convenience and brevity of description, specific working procedures of the above-described systems, apparatuses and units may refer to corresponding procedures in the foregoing method embodiments, which are not repeated herein.

The apparatus embodiments described above are merely illustrative, wherein the elements illustrated as separate elements may or may not be physically separate, and the elements shown as elements may or may not be physical elements, may be located in one place, or may be distributed over a plurality of network elements. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment. Those of ordinary skill in the art will understand and implement the present invention without undue burden.

From the above description of the embodiments, it will be apparent to those skilled in the art that the embodiments may be implemented by means of software plus necessary general hardware platforms, or of course may be implemented by means of hardware. Based on this understanding, the foregoing technical solution may be embodied essentially or in a part contributing to the prior art in the form of a software product, which may be stored in a computer readable storage medium, such as ROM/RAM, a magnetic disk, an optical disk, etc., including several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute the method described in the respective embodiments or some parts of the embodiments.

It should be noted that the above-mentioned embodiments are merely for illustrating the technical solution of the present application, and not for limiting the same, and although the present application has been described in detail with reference to the above-mentioned embodiments, it should be understood by those skilled in the art that the technical solution described in the above-mentioned embodiments may be modified or some technical features may be equivalently replaced, and these modifications or substitutions do not make the essence of the corresponding technical solution deviate from the spirit and scope of the technical solution of the embodiments of the present application.

Claims

1. A method for quickly identifying a specific drone model based on a radio frequency fingerprint library, comprising:

Receive the radio frequency signal generated by the target UAV during flight, and perform dynamic noise reduction processing on the radio frequency signal to generate a noise-reduced radio frequency signal;

Separate the combined data containing signal strength fluctuation pattern, phase hopping frequency and modulation distortion characteristics from the de-noised RF signal, and match the combined data with the pre-built RF fingerprint library layer by layer to generate a matching result;

Extracting model-related parameters of the target UAV from the matching results, the model-related parameters including the signal propagation delay difference corresponding to the reflection path of a specific obstacle and the frequency domain distortion correlation corresponding to the distribution of the electromagnetic interference source;

An identification decision including a real-time environment adaptation factor is generated based on the model association parameters, and a target UAV model result corresponding to the identification decision is output.

2. The method according to claim 1, wherein receiving a radio frequency signal generated by a target UAV during flight and dynamically performing noise reduction processing on the radio frequency signal to generate a noise-reduced radio frequency signal comprises:

The original radio frequency signal of the target UAV is captured in real time by a multi-channel signal acquisition device deployed in the monitoring area. The original radio frequency signal contains noise components caused by electromagnetic interference sources and obstacle reflections in the current scene;

Monitoring environmental parameters in the original radio frequency signal, wherein the environmental parameters include real-time interference intensity distribution and obstacle density distribution;

Selecting a signal filtering intensity level according to the environmental parameters; if the real-time interference intensity is higher than a set range and the obstacle density is higher than a set threshold, performing layer-by-layer noise suppression on the original radio frequency signal using multi-stage series filtering; and if the real-time interference intensity is lower than the set range, performing noise suppression on the original radio frequency signal using single-stage filtering to obtain a filtered radio frequency signal;

The filtered RF signal is processed for signal integrity restoration to generate a noise-reduced RF signal.

3. The method according to claim 1, characterized in that the method comprises separating combined data including signal strength fluctuation patterns, phase hop frequency, and modulation distortion characteristics from the noise-reduced RF signal, and matching the combined data layer by layer with a pre-built RF fingerprint library to generate a matching result, including:

Dividing the denoised radio frequency signal into multiple signal segments according to a preset time window, and extracting the fluctuation pattern of signal strength as it changes with flight altitude, the frequency distribution of phase jump intervals, and the position characteristics of modulation waveform distortion from each signal segment;

The fluctuation pattern, frequency distribution and location characteristics are integrated into combined data according to a preset weight ratio;

Calculate similarity between the combined data and the first-layer signal strength fluctuation pattern benchmark set stored in the radio frequency fingerprint library, and select a candidate model set whose similarity is higher than a first threshold;

The frequency distribution and position features in the combined data are jointly matched with the second-layer phase jump frequency reference set and modulation distortion reference set corresponding to the candidate model set to generate a matching result including the candidate model matching degree ranking.

4. The method according to claim 1, characterized in that model-related parameters of the target drone are extracted from the matching results, wherein the model-related parameters include the signal propagation delay difference corresponding to the reflection path of a specific obstacle and the frequency domain distortion correlation corresponding to the distribution of electromagnetic interference sources, including:

Sorting the candidate models in the matching results according to their matching scores, extracting the radio frequency fingerprint reference data corresponding to the candidate model with the highest ranking;

Obtaining a signal propagation delay difference parameter associated with obstacle distribution in a current scene from the radio frequency fingerprint reference data;

Acquire a frequency domain distortion correlation parameter associated with the electromagnetic interference source distribution from the radio frequency fingerprint reference data;

The signal propagation delay difference parameter and the frequency domain distortion correlation parameter are integrated into the model correlation parameter of the target UAV.

5. The method according to claim 1, wherein generating an identification decision including a real-time environment adaptation factor based on the model-related parameters and outputting a target UAV model result corresponding to the identification decision comprises:

Obtain real-time tracking data of the dynamic changes in the position of obstacles in the current scene and real-time fluctuation data of the intensity of electromagnetic interference sources;

Adjusting the weight coefficient of the signal propagation delay difference parameter according to the real-time tracking data, and adjusting the weight coefficient of the frequency domain distortion correlation parameter according to the real-time fluctuation data;

Input the adjusted signal propagation delay difference parameter and the adjusted frequency domain distortion correlation parameter into the environmental adaptation factor calculation model to generate a real-time environmental adaptation factor for correcting the matching degree of candidate models in the radio frequency fingerprint library;

Dynamically compensating the matching degree of candidate models in the matching results according to the real-time environment adaptation factor to generate an identification decision including a final matching model;

The model identifier of the final matching model in the identification decision is checked for consistency with the model feature description pre-stored in the radio frequency fingerprint library. When the consistency verification passes, the final matching model is output as the target drone model result.

6. The method according to claim 5, wherein adjusting the weight coefficient of the signal propagation delay difference parameter according to the real-time tracking data and adjusting the weight coefficient of the frequency domain distortion correlation parameter according to the real-time fluctuation data comprises:

Dynamically adjusting a weight coefficient of a signal propagation delay difference parameter based on a dynamic rate of change of an obstacle position in the real-time tracking data; when the dynamic rate exceeds a preset rate threshold, linearly enhancing the weight coefficient based on a ratio of the dynamic rate to the preset rate threshold to generate an adjusted delay weight coefficient;

According to the fluctuation amplitude of the electromagnetic interference source intensity in the real-time fluctuation data, the weight coefficient of the frequency domain distortion correlation parameter is dynamically adjusted. When the fluctuation amplitude exceeds the preset amplitude threshold, the weight coefficient is increased in segments according to the difference between the fluctuation amplitude and the amplitude threshold to generate an adjusted frequency domain weight coefficient.

7. The method according to claim 5, wherein dynamically compensating the matching degree of candidate models in the matching results according to the real-time environment adaptation factor to generate a recognition decision including a final matching model comprises:

Extracting an initial matching degree of the candidate model set from the matching results, and multiplying the delay correction factor by the signal propagation delay difference parameter corresponding to each candidate model in the initial matching degree item by item to generate a matching degree after delay compensation;

Multiplying the frequency domain correction factor by the frequency domain distortion correlation parameter corresponding to each candidate model in the matching degree after delay compensation item by item to generate the matching degree after frequency domain compensation;

The matching degree after delay compensation and the matching degree after frequency domain compensation are combined according to a preset ratio to generate a comprehensive compensation matching degree for each candidate model;

The candidate model set is sorted in descending order according to the comprehensive compensation matching degree, the candidate model with the highest ranking is selected as the final matching model, and an identification decision including the final matching model is generated.

8. A system for rapid identification of specific drone models based on a radio frequency fingerprint library, comprising:

A receiving module is used to receive the radio frequency signal generated by the target UAV during flight, and dynamically reduce the noise of the radio frequency signal to generate a noise-reduced radio frequency signal;

A matching module is used to separate the combined data containing the signal strength fluctuation pattern, phase hop frequency and modulation distortion characteristics from the noise-reduced RF signal, and match the combined data with the pre-built RF fingerprint library layer by layer to generate a matching result;

An extraction module is configured to extract model-related parameters of the target UAV from the matching results, wherein the model-related parameters include a signal propagation delay difference corresponding to a specific obstacle reflection path and a frequency domain distortion correlation corresponding to an electromagnetic interference source distribution;

The output module is used to generate an identification decision including a real-time environment adaptation factor based on the model association parameters, and output a target UAV model result corresponding to the identification decision.

9. A computing device, comprising a processing component and a storage component; the storage component stores one or more computer instructions; the one or more computer instructions are configured to be called and executed by the processing component to implement the method for rapidly identifying a specific drone model based on a radio frequency fingerprint library as described in any one of claims 1 to 7.

10. A computer storage medium, characterized in that it stores a computer program, which, when executed by a computer, implements the method for quickly identifying a specific drone model based on a radio frequency fingerprint library as described in any one of claims 1 to 7.