A method and apparatus for radio counter-jamming in the security defense of unmanned aerial vehicles (UAVs).

Abstract

This invention discloses a radio counter-jamming method and apparatus for unmanned aerial vehicle (UAV) security defense, relating to the field of radio counter-jamming. The method includes: constructing a UAV communication frequency band database and an interference source feature database; collecting spectrum data in the airspace to establish an airspace electromagnetic signal dataset; constructing an electromagnetic signal feature extraction model to obtain the time-domain features, frequency-hopping features, and modulation mode features of the electromagnetic signals; constructing an interference source identification model to eliminate electromagnetic signals from interference sources in the airspace electromagnetic signal dataset; constructing a UAV frequency band identification probability model to determine the probability that a frequency band in the airspace electromagnetic signal data belongs to the UAV communication frequency band; suppressing UAV communication signals using a radio counter-jamming device; comprehensively evaluating the radio counter-jamming effect and updating the database synchronously. The advantages of this invention are: it can effectively identify UAV frequency bands and eliminate interfering electromagnetic signals, and use a radio counter-jamming device to suppress UAV communication signals, ensuring the safety of the airspace.

Description

Technical Field

[0001] This invention relates to the field of radio counter-jamming, specifically to a radio counter-jamming method and apparatus for the security defense of unmanned aerial vehicles (UAVs). Background Technology

[0002] With the rapid iteration of consumer and industrial drone technology, its application scenarios have expanded to multiple fields such as logistics delivery, aerial photography and mapping, and security inspection. However, this has also brought severe airspace security challenges. Current drone communication technology is characterized by high-frequency frequency hopping, signal encryption, and multi-band multiplexing. For example, consumer drones generally use the 2.4GHz / 5.8GHz frequency band and support frequency hopping communication, while industrial drones introduce backup links such as 4G / 5G cellular networks and satellite communication. Some military drones even adopt encrypted frequency hopping protocols and anti-interference designs. These technological upgrades make traditional countermeasures based on fixed frequency band suppression ineffective, and there is an urgent need to build an intelligent defense system adapted to complex electromagnetic environments. In addition, the malicious deployment of human interference sources further exacerbates the difficulty of defense. Criminals deliberately confuse the signal characteristics of drones by placing fake base stations, portable jammers, and other devices, causing existing countermeasure systems to frequently misjudge. This can result in interference with legitimate equipment or, in severe cases, allow target drones to escape monitoring. Against this backdrop, defense solutions that integrate multi-dimensional signal recognition and probabilistic frequency band filtering have become a key direction for technological breakthroughs.

[0003] Traditional radio countermeasures equipment relies on a pre-set fixed frequency band library, which is insufficient for responding to high-frequency frequency-hopping drones. Judging by single frequency band characteristics or suppressing the entire frequency band can easily lead to the drone's interference communication signals being difficult to accurately capture and control, resulting in missed detections or wasted resources. Especially in the presence of interference sources, the effectiveness of radio countermeasures against drones will be greatly reduced, thus posing a risk to airspace security. Summary of the Invention

[0004] To address the aforementioned technical problems, this paper provides a radio counter-jamming method and device for drone security defense. This technical solution solves the problem mentioned in the background art that the response speed to high-frequency frequency-hopping drones is insufficient. Judging by single-band characteristics or suppressing the entire frequency band can easily lead to the drone's interference communication signals being difficult to accurately capture and control, resulting in missed detection or wasted resources. Especially in the presence of interference sources, the effectiveness of radio countermeasures against drones will be greatly reduced, thus posing a risk to airspace security.

[0005] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0006] A method for radio counter-jamming in unmanned aerial vehicle (UAV) security defense includes:

[0007] Collect and record the commonly used communication frequency bands of existing UAVs and the signal characteristics of common frequency interference sources, and construct a UAV communication frequency band library and an interference source feature library respectively;

[0008] Deploy ultra-wideband radio frequency antenna devices in multiple directions to collect spectrum data in the airspace, transmit the data to the central processing center using a wired network, and establish an airspace electromagnetic signal dataset.

[0009] Based on the spatial electromagnetic signal dataset, an electromagnetic signal feature extraction model is constructed to obtain the time-domain features, frequency hopping features, and modulation mode features of the electromagnetic signals.

[0010] Based on the interference source feature library, an interference source identification model is constructed to remove electromagnetic signals of interference sources from the spatial electromagnetic signal dataset.

[0011] Based on the UAV communication frequency band library and the Bayesian algorithm, a UAV frequency band identification probability model is constructed to determine the probability that the frequency band in the airspace electromagnetic signal data belongs to the UAV communication frequency band.

[0012] Based on the probability value of the central frequency band of the airspace electromagnetic signal data belonging to the UAV communication frequency band, the UAV communication signal is suppressed by the radio countermeasure device to ensure the safety of the airspace.

[0013] Based on feedback from airspace electromagnetic signal data, the effectiveness of radio countermeasures is comprehensively evaluated. When radio countermeasures fail, a strong magnetic excitation device is forcibly triggered, and the database is updated simultaneously.

[0014] Preferably, the step of constructing an electromagnetic signal feature extraction model based on the spatial electromagnetic signal dataset to obtain the time-domain features, frequency-hopping features, and modulation mode features of the electromagnetic signal specifically includes:

[0015] Based on the spatial electromagnetic signal dataset, the spatial electromagnetic signal data is converted into an electromagnetic signal time-frequency diagram using short-time Fourier transform.

[0016] Based on the time-frequency diagram of the electromagnetic signal, and using the sliding window algorithm, the signal frequency within a set time period is statistically analyzed to determine the frequency hopping period and rate of the electromagnetic signal.

[0017] Based on the time-frequency diagram of the electromagnetic signal, the modulation characteristics of the spatial electromagnetic signal are obtained using a convolutional neural network.

[0018] An electromagnetic signal feature extraction model is constructed to obtain the time-domain features, frequency hopping features, and modulation mode features of the electromagnetic signal.

[0019] Multi-dimensional electromagnetic signal feature vectors are constructed by feature concatenation, and the main features are extracted and feature dimensions are compressed based on the PCA algorithm.

[0020] Preferably, the step of constructing an interference source identification model based on the interference source feature library and removing electromagnetic signals of interference sources from the spatial electromagnetic signal dataset specifically includes:

[0021] The cosine similarity algorithm is used to calculate the similarity between multi-dimensional electromagnetic signal features and electromagnetic signal features in the interference source feature library;

[0022] Based on the time difference of arrival of signals collected by the multi-directional ultra-wideband radio frequency antenna device, the location of the electromagnetic signal source to be verified is obtained based on the TDOA algorithm.

[0023] Based on the multi-dimensional electromagnetic signal characteristics, and using the random forest algorithm, the confidence level that the electromagnetic signal is a UAV communication signal is obtained.

[0024] Based on the similarity value, the positional deviation of the electromagnetic signal source, and the confidence value, and using a linear weighting formula, it is determined whether the electromagnetic signal is an interference source. If it is, the electromagnetic signal of the interference source is removed from the airspace electromagnetic signal dataset; otherwise, it is determined to be an electromagnetic signal of a UAV.

[0025] To ensure the accuracy of interference source identification, a sliding window verification mechanism is set up, which means that the source is only removed after being identified as an interference source three times in a row, and the interference source data is retained for backtracking analysis of misjudgments.

[0026] Preferably, the step of constructing a UAV frequency band identification probability model based on the UAV communication frequency band library and the Bayesian algorithm to determine the probability that the frequency band in the airspace electromagnetic signal data belongs to the UAV communication frequency band specifically includes:

[0027] Based on the multi-dimensional electromagnetic signal characteristics and combined with the total probability formula, we obtain the evidence in the Bayesian algorithm, namely, the marginal probability of the occurrence of multi-dimensional electromagnetic signal characteristic data. By utilizing its normalization property, we ensure that the output probability value is in the [0,1] interval.

[0028] Based on the UAV communication frequency band library, interference source feature library and historical UAV identification data, the conditional probability of the occurrence of multi-dimensional electromagnetic signal features extracted by PCA is obtained by Gaussian distribution estimation and statistics under the conditions that the frequency band is a UAV communication frequency band and a non-UAV communication frequency band.

[0029] based on The criteria transform expert rules into numerical constraints to obtain the initial probability of this frequency band in UAV communication.

[0030] Based on the Bayesian algorithm, a probabilistic model for UAV frequency band identification is constructed to determine the probability that the frequency band in the airspace electromagnetic signal data belongs to the UAV communication frequency band.

[0031] Based on the collected multi-dimensional electromagnetic signal feature data, the probability values ​​of each frequency band are recalculated and updated at fixed time intervals to generate a dynamic probability frequency band list.

[0032] Preferably, the step of comprehensively evaluating the effectiveness of radio countermeasures based on feedback from spatial electromagnetic signal data, and forcibly triggering a strong magnetic excitation device and synchronously updating the database when radio countermeasures fail, specifically includes:

[0033] Based on the interference source location identification in the interference source identification model, and according to the feedback of the airspace electromagnetic signal data, it is determined whether the location of the electromagnetic signal source identified as the UAV is greater than the location deviation threshold.

[0034] Based on the feedback of airspace electromagnetic signal data, calculate whether the modulation disorder of the electromagnetic signal source identified as a UAV is greater than the disorder threshold.

[0035] Based on the feedback of airspace electromagnetic signal data, a dynamic map of the location of the electromagnetic signal source identified as the UAV, a time-frequency map of the electromagnetic signal, and a trend map of the period and rate of the frequency hopping signal are generated.

[0036] If both position deviation and disorder exceed the corresponding threshold, it is determined that a countermeasure effect has been generated. If neither position deviation nor disorder exceeds the corresponding threshold, it indicates that the radio countermeasure has failed. If one exceeds the threshold and the other does not, it is determined whether a countermeasure effect has been generated by combining the position dynamic diagram, the electromagnetic signal time-frequency diagram, and the period and rate trend diagram of the frequency hopping signal with expert experience.

[0037] If the radio countermeasures are deemed to have failed, the strong magnetic excitation device will be forcibly triggered to violently damage the drone's communication and control capabilities, as well as its internal electronic components and circuit systems.

[0038] Based on the results of radio countermeasures and the assessment results of expert experience, the UAV communication frequency band library and the interference source feature library are recorded and updated simultaneously.

[0039] Furthermore, this solution proposes a radio counter-jamming device for UAV security defense, used to implement the radio counter-jamming method for UAV security defense as described above, including:

[0040] The library module is used to collect and record the commonly used communication frequency bands of existing UAVs and the signal characteristics of common frequency interference sources, and to construct a UAV communication frequency band library and an interference source feature library, respectively.

[0041] The airspace monitoring module is used to deploy ultra-wideband radio frequency antenna devices in multiple directions, collect spectrum data in the airspace, transmit the data to the central processing center via a wired network, and establish an airspace electromagnetic signal dataset.

[0042] The feature extraction module is used to construct an electromagnetic signal feature extraction model based on the spatial electromagnetic signal dataset, and to obtain the time-domain features, frequency hopping features and modulation mode features of the electromagnetic signal respectively.

[0043] The interference signal removal module is used to construct an interference source identification model based on the interference source feature library and remove electromagnetic signals of interference sources from the spatial electromagnetic signal data.

[0044] The countermeasure and feedback module is used to construct a UAV frequency band identification probability model based on a UAV communication frequency band library and a Bayesian algorithm to determine the probability that a frequency band in the airspace electromagnetic signal data belongs to the UAV communication frequency band; based on the probability value of the frequency band in the airspace electromagnetic signal data belonging to the UAV communication frequency band, it suppresses the UAV communication signal through a radio countermeasure device to ensure the safety of the airspace; based on the feedback of the airspace electromagnetic signal data, it comprehensively evaluates the radio countermeasure effect, and when the radio countermeasure fails, it forcibly triggers a strong magnetic excitation device and updates the database synchronously.

[0045] Preferably, the countermeasure and feedback module specifically includes:

[0046] The frequency band probability unit is used to construct a UAV frequency band identification probability model based on the UAV communication frequency band library and the Bayesian algorithm to determine the probability that the frequency band in the airspace electromagnetic signal data belongs to the UAV communication frequency band.

[0047] A radio countermeasure unit is used to suppress UAV communication signals and ensure the safety of the airspace by using a radio countermeasure device based on the probability value that the central frequency band of the airspace electromagnetic signal data belongs to the UAV communication frequency band.

[0048] The countermeasure effect feedback unit is used to comprehensively evaluate the radio countermeasure effect based on the feedback of spatial electromagnetic signal data, and to forcibly trigger the strong magnetic excitation device and update the database synchronously when the radio countermeasure fails.

[0049] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0050] This invention provides a radio counter-jamming method and device for UAV security defense. Through an electromagnetic signal feature extraction model, an interference source identification model, and a UAV frequency band identification probability model, it achieves comprehensive UAV frequency band identification, from extracting the time-domain features, frequency-hopping features, and modulation mode features of electromagnetic signals, to eliminating electromagnetic signals from interference sources in the airspace electromagnetic signal data, and then determining the probability that the frequency band in the airspace electromagnetic signal data belongs to the UAV communication frequency band. This effectively filters interfering electromagnetic signals and selects the probability interval of UAV frequency bands. The radio counter-jamming device then effectively suppresses the frequency bands within the probability interval of the UAV frequency band. This effectively solves the problem that single-band feature judgment or full-band suppression can easily lead to the difficulty in accurately capturing and controlling UAV interference communication signals, thus improving the targeting of UAV communication frequency bands. Furthermore, to ensure airspace security, this invention proposes a feedback mechanism for radio counter-jamming, used to comprehensively evaluate the effectiveness of radio counter-jamming. When radio counter-jamming fails, a strong magnetic excitation device is forcibly triggered, and the database is updated synchronously, thereby further ensuring airspace security. Attached Figure Description

[0051] Figure 1 This is a flowchart of a radio counter-jamming method for drone security defense according to the present invention;

[0052] Figure 2 The flowchart of the present invention describes how to construct an electromagnetic signal feature extraction model based on a spatial electromagnetic signal dataset to obtain the time-domain features, frequency hopping features, and modulation mode features of electromagnetic signals.

[0053] Figure 3 The present invention provides an electromagnetic signal flowchart for constructing an interference source identification model based on an interference source feature library and removing interference sources from a spatial electromagnetic signal dataset.

[0054] Figure 4 The present invention provides a flowchart illustrating the probability of determining whether a frequency band in airspace electromagnetic signal data belongs to the UAV communication frequency band by constructing a UAV frequency band identification probability model based on a UAV communication frequency band library and a Bayesian algorithm. Detailed Implementation

[0055] The following description is intended to disclose the invention and enable those skilled in the art to implement it. The preferred embodiments described below are merely examples, and other obvious variations will occur to those skilled in the art.

[0056] Reference Figure 1 As shown, a radio counter-jamming method for drone security defense includes:

[0057] Collect and record the commonly used communication frequency bands of existing UAVs and the signal characteristics of common frequency interference sources, and construct a UAV communication frequency band library and an interference source feature library respectively;

[0058] Deploy ultra-wideband radio frequency antenna devices in multiple directions to collect spectrum data in the airspace, transmit the data to the central processing center using a wired network, and establish an airspace electromagnetic signal dataset.

[0059] Based on the spatial electromagnetic signal dataset, an electromagnetic signal feature extraction model is constructed to obtain the time-domain features, frequency hopping features, and modulation mode features of the electromagnetic signals.

[0060] Based on the interference source feature library, an interference source identification model is constructed to remove electromagnetic signals of interference sources from the spatial electromagnetic signal dataset.

[0061] Based on the UAV communication frequency band library and the Bayesian algorithm, a UAV frequency band identification probability model is constructed to determine the probability that the frequency band in the airspace electromagnetic signal data belongs to the UAV communication frequency band.

[0062] Based on the probability value of the central frequency band of the airspace electromagnetic signal data belonging to the UAV communication frequency band, the UAV communication signal is suppressed by the radio countermeasure device to ensure the safety of the airspace.

[0063] Based on feedback from airspace electromagnetic signal data, the effectiveness of radio countermeasures is comprehensively evaluated. When radio countermeasures fail, a strong magnetic excitation device is forcibly triggered, and the database is updated simultaneously.

[0064] It can be explained that radio counter-jamming involves interfering with the communication links or navigation systems of drones, rendering them inoperable and thus ensuring the safety of specific areas or facilities. Its core is to interfere with the drone's communication links or navigation systems through electromagnetic signals. For example, it can transmit strong interference signals in the same frequency band as the 2.4GHz / 5.8GHz band between the drone and its remote controller, blocking command transmission; or it can send false coordinates to the drone by simulating GPS / BeiDou satellite signals, inducing it to deviate from its intended flight path. Therefore, accurately capturing the drone's communication frequency band is crucial for effective radio counter-jamming. This solution utilizes an electromagnetic signal feature extraction model, an interference source identification model, and drone... The UAV frequency band identification probability model achieves comprehensive UAV frequency band identification, from extracting the time-domain characteristics, frequency hopping characteristics, and modulation mode characteristics of electromagnetic signals, to eliminating electromagnetic signals from interference sources in the airspace electromagnetic signal data, and then determining the probability that the frequency band in the airspace electromagnetic signal data belongs to the UAV communication frequency band. It effectively filters interfering electromagnetic signals and selects the probability range of UAV frequency bands. Then, through radio countermeasures devices, it can effectively suppress the frequency bands in the probability range of UAV frequency bands. It can effectively solve the problem that single-band feature judgment or full-band suppression can easily lead to the difficulty in accurately capturing and controlling UAV interference communication signals, thereby improving the targeting of UAV communication frequency bands.

[0065] The collection and recording of commonly used communication frequency bands and common frequency interference source signal characteristics of existing UAVs, and the construction of UAV communication frequency band databases and interference source feature databases respectively, specifically include:

[0066] Based on the existing types of drones, the commonly used communication frequency bands for different types of drones are obtained. The commonly used communication frequency bands include: civilian drone frequency bands, special purpose drone frequency bands, navigation and positioning frequency bands, and emerging communication frequency bands.

[0067] A UAV communication frequency band library was created using a MySQL relational database, based on commonly used communication frequency bands for different UAV types.

[0068] The database settings include frequency band range, frequency band purpose, common models used, signal modulation method, and frequency band signal characteristics;

[0069] Collect and record the signal characteristics of common frequency interference sources to construct an interference source feature library. The interference source signal characteristics include: fixed frequency interference sources, pulse interference sources, and mobile interference sources.

[0070] Using spectrum monitoring equipment, we regularly record and collect the characteristics of UAV communication frequency bands and interference sources that are not included in the UAV communication frequency band database, in order to update the database.

[0071] Explained by this, by acquiring common UAV communication frequency bands and the signal characteristics of interference sources, comparative analysis can be used to effectively identify and differentiate them, thereby enabling targeted radio counter-interference. Therefore, this solution collects and records common UAV communication frequency bands and the signal characteristics of interference sources to construct a UAV communication frequency band library and an interference source feature library, thus providing data support for UAV frequency band identification and interference source elimination. Specifically, for fixed-frequency interference sources, such as radio stations (AM broadcast 526.5-1606.5kHz, FM broadcast 87-108MHz) and communication base stations (specific frequency bands of 2G, 3G, 4G, and 5G base stations), their transmission frequency, signal strength, and modulation method are recorded. The system identifies and stores these characteristics in an interference source feature library. For pulse-type interference sources, such as radar signals (common radar frequency bands include S-band 2-4GHz and C-band 4-8GHz) and electrical spark interference, the system extracts their pulse width, pulse repetition frequency, and signal spectrum characteristics and incorporates them into the interference source feature library. For mobile interference sources, such as vehicle-mounted jamming devices and portable jammers, whose signal characteristics have a certain degree of dynamic change, the system collects the signal characteristics of these interference sources at different moving speeds and different transmission powers, including the change law of signal strength over time and frequency drift, and incorporates them into the interference source feature library. This improves the comprehensiveness and diversity of the interference source library data, and the library data is updated based on the subsequent countermeasure results.

[0072] The multi-directional deployment of ultra-wideband radio frequency antenna devices, which collects spectral data in the airspace, transmits the data to the central processing center via a wired network, and establishes an airspace electromagnetic signal dataset, specifically includes:

[0073] An ultra-wideband radio frequency antenna was selected to ensure that it could cover the frequency range of 300MHz-6GHz, so as to comprehensively monitor the frequency band used by the drone;

[0074] Around the protected area, multiple spectrum monitoring nodes are deployed at certain intervals. Each node is equipped with an independent ultra-wideband radio frequency antenna and a preprocessing unit, and converts the collected raw signals into digital signals.

[0075] Based on the Kalman filter algorithm, the collected spectrum data is filtered and denoised. At the same time, the min-max normalization formula is used to normalize the collected spectrum data to eliminate the influence of data units.

[0076] Each monitoring node transmits pre-processed data to the central processing center via a wired network, thereby aggregating and integrating the data from different monitoring nodes to construct a spatial electromagnetic signal dataset.

[0077] This can be explained by the fact that deploying ultra-wideband radio frequency antennas across multiple nodes effectively covers the frequency range of 300MHz-6GHz, enabling comprehensive capture of signals from civilian and special-purpose drones and emerging communication frequency bands. This avoids missed detections due to incomplete frequency band coverage. Secondly, transmitting pre-processed data to the central processing center via a wired network avoids wireless transmission occupying the same frequency band resources used by drones for communication, preventing frequency band overlap between the transmitted signal and the monitored drone signal or interference source signal, which would introduce additional electromagnetic interference, affect the purity of the original spectrum data, and interfere with the accuracy of subsequent feature extraction and identification. At the same time, it can effectively meet the massive amount of spectrum data collected by ultra-wideband radio frequency antennas, ensuring stability and controllable latency, which meets the real-time data processing needs of high-frequency frequency-hopping drones. Furthermore, wired transmission, through physical link isolation and combined with encryption protocols, significantly reduces the risk of eavesdropping and tampering, ensuring data integrity, which is crucial for airspace security.

[0078] Reference Figure 2 As shown, the step of constructing an electromagnetic signal feature extraction model based on the spatial electromagnetic signal dataset to obtain the time-domain features, frequency-hopping features, and modulation mode features of the electromagnetic signal specifically includes:

[0079] Based on the spatial electromagnetic signal dataset, the spatial electromagnetic signal data is converted into an electromagnetic signal time-frequency diagram using short-time Fourier transform.

[0080] Based on the time-frequency diagram of the electromagnetic signal, and using the sliding window algorithm, the signal frequency within a set time period is statistically analyzed to determine the frequency hopping period and rate of the electromagnetic signal.

[0081] Based on the time-frequency diagram of the electromagnetic signal, the modulation characteristics of the spatial electromagnetic signal are obtained using a convolutional neural network.

[0082] An electromagnetic signal feature extraction model is constructed to obtain the time-domain features, frequency hopping features, and modulation mode features of the electromagnetic signal.

[0083] Multi-dimensional electromagnetic signal feature vectors are constructed by feature concatenation, and the main features are extracted and feature dimensions are compressed based on the PCA algorithm.

[0084] It can be explained that the key to identifying UAV communication frequency bands and eliminating interference sources lies in the feature matching of electromagnetic signals. Therefore, when identifying UAV communication frequency bands and eliminating interference sources, it is necessary to extract the electromagnetic signal features in advance to ensure that the electromagnetic signal features can be effectively matched. Therefore, this solution extracts the time domain features, frequency hopping features, and modulation mode features of electromagnetic signals respectively, and extracts the main features and compresses the feature dimensions based on the PCA algorithm, thereby improving the feature matching of electromagnetic signals. Specifically, the conversion of spatial electromagnetic signal data into electromagnetic signal time-frequency diagram based on short-time Fourier transform includes: dividing the original signal of the spatial electromagnetic signal into multiple short-time overlapping windows.

[0085] Perform a Fourier transform on the signal within each window to obtain the complex spectrum;

[0086] Arrange the complex spectra of all windows in chronological order to form a complex spectrum matrix;

[0087] The modulus of the complex spectrum matrix is ​​taken, and based on the logarithmic formula, it is converted into a logarithmic scale to obtain the time-frequency diagram of the logarithmic amplitude spectrum of the electromagnetic signal.

[0088] The short-time Fourier transform expression is as follows:

[0089]

[0090] In the formula, In time and frequency Complex spectral values ​​at that location, The original signal of the spatial electromagnetic signal. For Hanning window functions, Let the integral variable be time. The frequency of the electromagnetic signal in the spatial domain. The imaginary unit, The length of the window;

[0091] The logarithmic formula is converted into a logarithmic scale expression as follows:

[0092]

[0093] In the formula, In time and frequency The logarithmic scale of the complex spectrum values ​​is used to better align with human auditory perception characteristics or the need for dynamic range compression. As a correction term, to avoid logarithmic overflow, it is adjusted according to the dynamic range of the spatial electromagnetic signal;

[0094] The step of statistically analyzing the signal frequency within a set time period based on the sliding window algorithm to determine the frequency hopping period and rate of the electromagnetic signal specifically includes:

[0095] The logarithmic scale matrix of the complex spectrum values ​​is used as the input to the sliding window algorithm;

[0096] Based on the input, combined with the sliding window algorithm, the frequency energy distribution of the electromagnetic signal in the spatial domain is obtained, and the maximum local frequency energy is selected as the dominant frequency point.

[0097] Determine whether the same dominant frequency point appears consecutively in multiple windows (the number of windows can be set to 3). If yes, it is determined to be the current stationary frequency band; otherwise, it is determined to be a non-current stationary frequency band.

[0098] Determine whether the absolute value of the frequency energy difference between the current camping frequency band and the previous current camping frequency band is greater than the hopping threshold. If yes, it is determined to be a frequency band hopping; otherwise, it is determined to be a frequency band not hopping.

[0099] Among them, the jump threshold is a value determined based on historical monitoring data, normal distribution, and expert experience.

[0100] Based on the identification results of frequency band hopping, calculate the frequency hopping period and rate of the electromagnetic signal;

[0101] The expression for the frequency energy distribution of the spatial electromagnetic signal is:

[0102]

[0103] In the formula, For time Starting point, width is Within the time window, the frequency is The frequency energy value of the electromagnetic signal in the spatial domain at that location. The length of the sliding window. The start time;

[0104] The frequency hopping period expression of the electromagnetic signal is:

[0105]

[0106] In the formula, This is a median estimate of the frequency hopping period of the electromagnetic signal. This represents the total number of frequency hopping operations for electromagnetic signals. For the first The timestamp of the frequency hop. For the first The timestamp of the frequency hop. This is the median function, which takes the median of the set of all time intervals within the parentheses.

[0107] The frequency hopping rate expression for the electromagnetic signal is:

[0108]

[0109] In the formula, This represents the frequency hopping rate of the electromagnetic signal.

[0110] Reference Figure 3 As shown, the step of constructing an interference source identification model based on the interference source feature library and removing electromagnetic signals from the spatial electromagnetic signal dataset that represent interference sources specifically includes:

[0111] The cosine similarity algorithm is used to calculate the similarity between multi-dimensional electromagnetic signal features and electromagnetic signal features in the interference source feature library;

[0112] Based on the time difference of arrival of signals collected by the multi-directional ultra-wideband radio frequency antenna device, the location of the electromagnetic signal source to be verified is obtained based on the TDOA algorithm.

[0113] Based on the multi-dimensional electromagnetic signal characteristics, and using the random forest algorithm, the confidence level that the electromagnetic signal is a UAV communication signal is obtained.

[0114] Based on the similarity value, the positional deviation of the electromagnetic signal source, and the confidence value, and using a linear weighting formula, it is determined whether the electromagnetic signal is an interference source. If it is, the electromagnetic signal of the interference source is removed from the airspace electromagnetic signal dataset; otherwise, it is determined to be an electromagnetic signal of a UAV.

[0115] To ensure the accuracy of interference source identification, a sliding window verification mechanism is set up, which means that the source is only removed after being identified as an interference source three times in a row, and the interference source data is retained for backtracking analysis of misjudgments.

[0116] This solution involves collecting and recording commonly used UAV frequency band data, comparing it with airspace spectrum data, and using a Bayesian algorithm to estimate the probability of UAV frequency bands for accurate matching. A prerequisite for this is effectively removing electromagnetic signals from interference sources, especially those maliciously deployed by individuals. Therefore, this solution uses cosine similarity, TDOA, and random forest algorithms. For fixed-deployment fake base stations, TDOA positioning captures their unchanging location. The random forest algorithm weights the motion characteristics of electromagnetic signals, identifying interference sources based on differences in location dynamics. For the motion characteristics of interference sources, cosine similarity calculations reveal subtle deviations in signal power stability and frequency hopping period uniformity. A sliding window verification (three consecutive checks) is used to improve the ability to eliminate interference sources.

[0117] The process of obtaining the confidence level that the electromagnetic signal is a UAV communication signal based on the random forest algorithm specifically includes:

[0118] Based on the UAV communication frequency band library, it is labeled as positive sample data of electromagnetic signal feature vectors of UAV communication signals;

[0119] Based on the interference source feature library, it is labeled as negative sample data of interference source and noise signal feature vector;

[0120] Based on the random forest algorithm, the probability of the electromagnetic signal being a UAV communication signal is output through a voting mechanism, i.e., the confidence level.

[0121] The confidence expression is as follows:

[0122]

[0123] In the formula, The confidence level of the electromagnetic signal is that it is the communication signal of the UAV. The total number of decision trees in the random forest. For the first Each decision tree predicts the category of the drone. This is an indicator function; its value is 1 when the condition within the parentheses is true, and 0 otherwise.

[0124] The formula based on the similarity value, the positional deviation of the electromagnetic signal source, and the confidence level is as follows:

[0125]

[0126] In the formula, The electromagnetic signal is a comprehensive evaluation value for interference signal sources. , , These are the linear regression coefficients. For bias terms, The similarity value is... The positional deviation of the electromagnetic signal source is given by, where, , , , It was obtained by solving historical data using the least squares method. It is based on Euclidean distance to obtain the position deviation value from the coordinate value;

[0127] During calculation, the parameters need to be normalized using the min-max method to eliminate the influence of data dimensions. Specifically, the method for determining whether an electromagnetic signal is an interference source is as follows: based on historical data... The criteria, combined with expert experience, set an initial interference threshold. By comparing the comprehensive evaluation value with the initial interference threshold, if it exceeds the initial interference threshold, it is determined to be an interference source; otherwise, it is considered a non-interference source.

[0128] Reference Figure 4 As shown, the step of constructing a UAV frequency band identification probability model based on the UAV communication frequency band library and the Bayesian algorithm to determine the probability that the frequency band in the airspace electromagnetic signal data belongs to the UAV communication frequency band specifically includes:

[0129] Based on the multi-dimensional electromagnetic signal characteristics and combined with the total probability formula, we obtain the evidence in the Bayesian algorithm, namely, the marginal probability of the occurrence of multi-dimensional electromagnetic signal characteristic data. By utilizing its normalization property, we ensure that the output probability value is in the [0,1] interval.

[0130] Based on the UAV communication frequency band library, interference source feature library and historical UAV identification data, the conditional probability of the occurrence of multi-dimensional electromagnetic signal features extracted by PCA is obtained by Gaussian distribution estimation and statistics under the conditions that the frequency band is a UAV communication frequency band and a non-UAV communication frequency band.

[0131] based on The criteria transform expert rules into numerical constraints to obtain the initial probability of this frequency band in UAV communication.

[0132] Based on the Bayesian algorithm, a probabilistic model for UAV frequency band identification is constructed to determine the probability that the frequency band in the airspace electromagnetic signal data belongs to the UAV communication frequency band.

[0133] Based on the collected multi-dimensional electromagnetic signal feature data, the probability values ​​of each frequency band are recalculated and updated at fixed time intervals to generate a dynamic probability frequency band list.

[0134] This solution utilizes a Bayesian algorithm to dynamically quantify the probability that each frequency band in the airspace electromagnetic signal belongs to the UAV communication band. It transforms the traditional hard decision (yes / no UAV frequency band) into continuous probability values ​​([0,1]), improving decision-making flexibility. Furthermore, it updates and integrates the UAV communication band library, interference source feature library, and historical data, reducing the false positive rate. This provides real-time, interpretable data support for spectrum monitoring, interference countermeasures, and resource allocation, adapting it to time-varying scenarios such as UAV frequency hopping and new types of interference.

[0135] The evidence expression obtained by combining the law of total probability in the Bayesian algorithm is as follows:

[0136]

[0137] In the formula, The marginal probability of the occurrence of multi-dimensional electromagnetic signal features extracted by PCA, i.e., evidence. This is the multi-dimensional electromagnetic signal feature vector extracted by PCA. This represents the conditional probability of the occurrence of multi-dimensional electromagnetic signal features extracted by PCA under the given frequency band conditions of UAV communication. This represents the conditional probability of the occurrence of multi-dimensional electromagnetic signal features extracted by PCA, given that the frequency band is not a UAV communication frequency band. This represents the initial probability that this frequency band is the communication frequency band for drones. Let be the initial probability that this frequency band is not a drone communication frequency band, where ;

[0138] The expression for the UAV frequency band identification probability model is as follows:

[0139]

[0140] In the formula, Given the evidence, the probability that this frequency band is a communication frequency band for drones is calculated.

[0141] The method of suppressing UAV communication signals and ensuring airspace security by using a radio countermeasure device, based on the probability value of the central frequency band in the airspace electromagnetic signal data belonging to the UAV communication frequency band, specifically includes:

[0142] The radio countermeasure device is based on software-defined radio technology, which supports flexible switching of operating frequency bands within the 300MHz-6GHz frequency band range and can simultaneously transmit 1-4 sets of countermeasure signals in different frequency bands.

[0143] Based on the data in the UAV communication frequency band library, pre-configure corresponding countermeasure signals for different frequency bands for each radio countermeasure device;

[0144] The radio countermeasure device is equipped with an adaptive power amplifier, which can automatically adjust the transmission power of the countermeasure signal according to the distance and signal strength of the target drone;

[0145] Based on the probability values ​​of the frequency bands in the airspace electromagnetic signal data belonging to the UAV communication frequency bands, the frequency band types with probability values ​​exceeding the preset value are filtered out.

[0146] Based on the selected frequency band type, the corresponding radio countermeasure device is activated to actively transmit countermeasure signals to interfere with the drone's communication signals.

[0147] This can be explained by using a drone frequency band identification probability model to filter out frequency band types with probability values ​​exceeding a preset value. This can be one or multiple frequency band types. By deploying multiple sets of radio countermeasure devices, each set of radio countermeasure devices simultaneously transmits 1-4 sets of countermeasure signals in different frequency bands, thereby eliminating frequency band interference with low probability and improving the targeting and comprehensiveness of radio countermeasures. The preset value is based on historical data, plotting ROC curves on the past 1000 sets of samples, and selecting the critical probability value that makes the TPR reach 95% and the FPR below 5%.

[0148] The process of comprehensively evaluating the effectiveness of radio countermeasures based on feedback from spatial electromagnetic signal data, and forcibly triggering a strong magnetic excitation device and synchronously updating the database when radio countermeasures fail, specifically includes:

[0149] Based on the interference source location identification in the interference source identification model, and according to the feedback of the airspace electromagnetic signal data, it is determined whether the location of the electromagnetic signal source identified as the UAV is greater than the location deviation threshold.

[0150] Based on the feedback of airspace electromagnetic signal data, calculate whether the modulation disorder of the electromagnetic signal source identified as a UAV is greater than the disorder threshold.

[0151] Based on the feedback of airspace electromagnetic signal data, a dynamic map of the location of the electromagnetic signal source identified as the UAV, a time-frequency map of the electromagnetic signal, and a trend map of the period and rate of the frequency hopping signal are generated.

[0152] If both position deviation and disorder exceed the corresponding threshold, it is determined that a countermeasure effect has been generated. If neither position deviation nor disorder exceeds the corresponding threshold, it indicates that the radio countermeasure has failed. If one exceeds the threshold and the other does not, it is determined whether a countermeasure effect has been generated by combining the position dynamic diagram, the electromagnetic signal time-frequency diagram, and the period and rate trend diagram of the frequency hopping signal with expert experience.

[0153] If the radio countermeasures are deemed to have failed, the strong magnetic excitation device will be forcibly triggered to violently damage the drone's communication and control capabilities, as well as its internal electronic components and circuit systems.

[0154] Based on the results of radio countermeasures and the assessment results of expert experience, the UAV communication frequency band library and the interference source feature library are recorded and updated simultaneously.

[0155] This solution uses positional deviation to determine if a drone has lost control and turbulence to determine if its communication signals have been effectively interfered with. When quantitative indicators conflict, such as signal turbulence without positional shift, visual data such as position dynamics and time-frequency graphs, combined with expert judgment of drone behavior patterns, improve the accuracy of countermeasure effectiveness assessment and reduce the risk of misjudging effectiveness or missing failure. Simultaneously, it uses the forced triggering of a strong magnetic field as a baseline, physically damaging the drone's communication and control capabilities, as well as its internal electronic components and circuitry, through a directional strong magnetic field, thus forcibly terminating its threat.

[0156] The disorder degree expression is:

[0157]

[0158] In the formula, The error vector amplitude, or disorder, is the electromagnetic signal source identified as the UAV. The number of samples used in the calculation. The first spatial electromagnetic signal acquired after countermeasure One in-phase component sample, The first spatial electromagnetic signal acquired after countermeasure orthogonal component samples , For a signal to reach its ideal state under a specific modulation scheme (such as QPSK, 16QAM), it must be composed of a set of ideal in-phase ( ) and orthogonal ( The signal quality is represented by component coordinates, which are matched from the UAV communication frequency band library and serve as a reference standard for evaluating the actual signal quality. The average power of the ideal symbol is obtained by calculating the sum of the squares of the ideal in-phase and quadrature components, i.e. This value serves as the reference power for calculating the EVM, used to normalize the difference between the actual and ideal signals. , It obtains the signal from a pre-stored ideal constellation template based on the modulation scheme detected in real time (such as QPSK, 16QAM), and the reversed measured signal is then processed. , A dynamic constellation diagram is generated, and the EVM is calculated by comparing the two. The constellation diagram is generated by the H / Q signals in the reversed spatial electromagnetic signal.

[0159] Among them, the position deviation threshold is combined with the UAV flight status and area protection level. The spatiotemporal characteristics of historical trajectories are analyzed through LSTM network, such as heading angle variance and acceleration extreme values, which are dynamically generated and updated online every 5 minutes. The modulation disorder threshold is based on the 3GPP standard baseline value, such as QPSK: 35% and 16QAM: 25%, with the addition of equipment performance correction terms (Δ=0.1×interference power / dBm). Combined with expert experience, the parameters are recalibrated weekly based on measured data.

[0160] Furthermore, based on the same inventive concept as the aforementioned radio counter-jamming method for drone security defense, this solution proposes a radio counter-jamming device for drone security defense, comprising:

[0161] The library module is used to collect and record the commonly used communication frequency bands of existing UAVs and the signal characteristics of common frequency interference sources, and to construct a UAV communication frequency band library and an interference source feature library, respectively.

[0162] The airspace monitoring module is used to deploy ultra-wideband radio frequency antenna devices in multiple directions, collect spectrum data in the airspace, transmit the data to the central processing center via a wired network, and establish an airspace electromagnetic signal dataset.

[0163] The feature extraction module is used to construct an electromagnetic signal feature extraction model based on the spatial electromagnetic signal dataset, and to obtain the time-domain features, frequency hopping features and modulation mode features of the electromagnetic signal respectively.

[0164] The interference signal removal module is used to construct an interference source identification model based on the interference source feature library and remove electromagnetic signals of interference sources from the spatial electromagnetic signal data.

[0165] The countermeasure and feedback module is used to construct a UAV frequency band identification probability model based on a UAV communication frequency band library and a Bayesian algorithm to determine the probability that a frequency band in the airspace electromagnetic signal data belongs to the UAV communication frequency band; based on the probability value of the frequency band in the airspace electromagnetic signal data belonging to the UAV communication frequency band, it suppresses the UAV communication signal through a radio countermeasure device to ensure the safety of the airspace; based on the feedback of the airspace electromagnetic signal data, it comprehensively evaluates the radio countermeasure effect, and when the radio countermeasure fails, it forcibly triggers a strong magnetic excitation device and updates the database synchronously;

[0166] The countermeasure and feedback module specifically includes:

[0167] The frequency band probability unit is used to construct a UAV frequency band identification probability model based on the UAV communication frequency band library and the Bayesian algorithm to determine the probability that the frequency band in the airspace electromagnetic signal data belongs to the UAV communication frequency band.

[0168] A radio countermeasure unit is used to suppress UAV communication signals and ensure the safety of the airspace by using a radio countermeasure device based on the probability value that the central frequency band of the airspace electromagnetic signal data belongs to the UAV communication frequency band.

[0169] The countermeasure effect feedback unit is used to comprehensively evaluate the radio countermeasure effect based on the feedback of spatial electromagnetic signal data, and to forcibly trigger the strong magnetic excitation device and update the database synchronously when the radio countermeasure fails.

[0170] In summary, the advantages of this invention are: it can effectively identify the frequency band of UAVs and eliminate interfering electromagnetic signals, and use radio countermeasures devices to suppress UAV communication signals, thereby ensuring the safety of the airspace.

[0171] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the claimed invention. The scope of protection claimed by the appended claims and their equivalents is defined.

Claims

1. A method for radio counter-jamming in the security defense of unmanned aerial vehicles (UAVs), characterized in that, include: Collect and record the commonly used communication frequency bands of existing UAVs and the signal characteristics of common frequency interference sources, and construct a UAV communication frequency band library and an interference source feature library respectively; Deploy ultra-wideband radio frequency antenna devices in multiple directions to collect spectrum data in the airspace, transmit the data to the central processing center using a wired network, and establish an airspace electromagnetic signal dataset. Based on the spatial electromagnetic signal dataset, an electromagnetic signal feature extraction model is constructed to obtain the time-domain features, frequency hopping features, and modulation mode features of the electromagnetic signals. Based on the interference source feature library, an interference source identification model is constructed to remove electromagnetic signals of interference sources from the spatial electromagnetic signal dataset. Based on the UAV communication frequency band library and the Bayesian algorithm, a UAV frequency band identification probability model is constructed to determine the probability that the frequency band in the airspace electromagnetic signal data belongs to the UAV communication frequency band. Based on the probability value of the central frequency band of the airspace electromagnetic signal data belonging to the UAV communication frequency band, the UAV communication signal is suppressed by the radio countermeasure device to ensure the safety of the airspace. Based on the feedback of airspace electromagnetic signal data, the effectiveness of radio countermeasures is comprehensively evaluated, and when radio countermeasures fail, a strong magnetic excitation device is forcibly triggered and the database is updated simultaneously. The process of comprehensively evaluating the effectiveness of radio countermeasures based on feedback from spatial electromagnetic signal data, and forcibly triggering a strong magnetic excitation device and synchronously updating the database when radio countermeasures fail, specifically includes: Based on the interference source location identification in the interference source identification model, and according to the feedback of the airspace electromagnetic signal data, it is determined whether the location of the electromagnetic signal source identified as the UAV is greater than the location deviation threshold. Based on the feedback of airspace electromagnetic signal data, calculate whether the modulation disorder of the electromagnetic signal source identified as a UAV is greater than the disorder threshold. Based on the feedback of airspace electromagnetic signal data, a dynamic map of the location of the electromagnetic signal source identified as the UAV, a time-frequency map of the electromagnetic signal, and a trend map of the period and rate of the frequency hopping signal are generated. If both position deviation and disorder exceed the corresponding threshold, it is determined that a countermeasure effect has been generated. If neither position deviation nor disorder exceeds the corresponding threshold, it indicates that the radio countermeasure has failed. If one exceeds the threshold and the other does not, it is determined whether a countermeasure effect has been generated by combining the position dynamic diagram, the electromagnetic signal time-frequency diagram, and the period and rate trend diagram of the frequency hopping signal with expert experience. If the radio countermeasures are deemed to have failed, the strong magnetic excitation device will be forcibly triggered to violently damage the drone's communication and control capabilities, as well as its internal electronic components and circuit systems. Based on the results of radio countermeasures and the assessment results of expert experience, the UAV communication frequency band library and the interference source feature library are recorded and updated simultaneously.

2. The radio counter-jamming method for UAV security defense according to claim 1, characterized in that, The collection and recording of commonly used communication frequency bands and common frequency interference source signal characteristics of existing UAVs, and the construction of UAV communication frequency band databases and interference source feature databases respectively, specifically include: Based on the existing types of drones, the commonly used communication frequency bands for different types of drones are obtained. The commonly used communication frequency bands include: civilian drone frequency bands, special purpose drone frequency bands, navigation and positioning frequency bands, and emerging communication frequency bands. A UAV communication frequency band library was created using a MySQL relational database, based on commonly used communication frequency bands for different UAV types. The database settings include frequency band range, frequency band purpose, common models used, signal modulation method, and frequency band signal characteristics; Collect and record the signal characteristics of common frequency interference sources to construct an interference source feature library. The interference source signal characteristics include: fixed frequency interference sources, pulse interference sources, and mobile interference sources. Using spectrum monitoring equipment, we regularly record and collect the characteristics of UAV communication frequency bands and interference sources that are not included in the UAV communication frequency band database, in order to update the database.

3. The radio counter-jamming method for UAV security defense according to claim 2, characterized in that, The multi-directional deployment of ultra-wideband radio frequency antenna devices, which collects spectral data in the airspace, transmits the data to the central processing center via a wired network, and establishes an airspace electromagnetic signal dataset, specifically includes: An ultra-wideband radio frequency antenna was selected to ensure that it could cover the frequency range of 300MHz-6GHz, so as to comprehensively monitor the frequency band used by the drone; Around the protected area, multiple spectrum monitoring nodes are deployed at certain intervals. Each node is equipped with an independent ultra-wideband radio frequency antenna and a preprocessing unit, and converts the collected raw signals into digital signals. Based on the Kalman filter algorithm, the collected spectrum data is filtered and denoised. At the same time, the min-max normalization formula is used to normalize the collected spectrum data to eliminate the influence of data units. Each monitoring node transmits pre-processed data to the central processing center via a wired network, thereby aggregating and integrating the data from different monitoring nodes to construct a spatial electromagnetic signal dataset.

4. The radio counter-jamming method for UAV security defense according to claim 3, characterized in that, The step of constructing an electromagnetic signal feature extraction model based on a spatial electromagnetic signal dataset to obtain the time-domain features, frequency-hopping features, and modulation mode features of the electromagnetic signals specifically includes: Based on the spatial electromagnetic signal dataset, the spatial electromagnetic signal data is converted into an electromagnetic signal time-frequency diagram using short-time Fourier transform. Based on the time-frequency diagram of the electromagnetic signal, and using the sliding window algorithm, the signal frequency within a set time period is statistically analyzed to determine the frequency hopping period and rate of the electromagnetic signal. Based on the time-frequency diagram of the electromagnetic signal, the modulation characteristics of the spatial electromagnetic signal are obtained using a convolutional neural network. An electromagnetic signal feature extraction model is constructed to obtain the time-domain features, frequency hopping features, and modulation mode features of the electromagnetic signal. Multi-dimensional electromagnetic signal feature vectors are constructed by feature concatenation, and the main features are extracted and feature dimensions are compressed based on the PCA algorithm.

5. A radio counter-jamming method for UAV security defense according to claim 4, characterized in that, The step of constructing an interference source identification model based on the interference source feature library and removing electromagnetic signals from the spatial electromagnetic signal dataset that represent interference sources specifically includes: The cosine similarity algorithm is used to calculate the similarity between multi-dimensional electromagnetic signal features and electromagnetic signal features in the interference source feature library; Based on the time difference of arrival of signals collected by the multi-directional ultra-wideband radio frequency antenna device, the location of the electromagnetic signal source to be verified is obtained based on the TDOA algorithm. Based on the multi-dimensional electromagnetic signal characteristics, and using the random forest algorithm, the confidence level that the electromagnetic signal is a UAV communication signal is obtained. Based on the similarity value, the positional deviation of the electromagnetic signal source, and the confidence value, and using a linear weighting formula, it is determined whether the electromagnetic signal is an interference source. If it is, the electromagnetic signal of the interference source is removed from the airspace electromagnetic signal dataset; otherwise, it is determined to be an electromagnetic signal of a UAV. To ensure the accuracy of interference source identification, a sliding window verification mechanism is set up, which means that the source is only removed after being identified as an interference source three times in a row, and the interference source data is retained for backtracking analysis of misjudgments.

6. A radio counter-jamming method for UAV security defense according to claim 5, characterized in that, The step of constructing a UAV frequency band identification probability model based on a UAV communication frequency band library and a Bayesian algorithm to determine the probability that a frequency band in airspace electromagnetic signal data belongs to a UAV communication frequency band specifically includes: Based on the multi-dimensional electromagnetic signal characteristics and combined with the total probability formula, we obtain the evidence in the Bayesian algorithm, namely, the marginal probability of the occurrence of multi-dimensional electromagnetic signal characteristic data. By utilizing its normalization property, we ensure that the output probability value is in the [0,1] interval. Based on the UAV communication frequency band library, interference source feature library and historical UAV identification data, the conditional probability of the occurrence of multi-dimensional electromagnetic signal features extracted by PCA is obtained by Gaussian distribution estimation and statistics under the conditions that the frequency band is a UAV communication frequency band and a non-UAV communication frequency band. based on The criteria transform expert rules into numerical constraints to obtain the initial probability of this frequency band in UAV communication. Based on the Bayesian algorithm, a probabilistic model for UAV frequency band identification is constructed to determine the probability that the frequency band in the airspace electromagnetic signal data belongs to the UAV communication frequency band. Based on the collected multi-dimensional electromagnetic signal feature data, the probability values ​​of each frequency band are recalculated and updated at fixed time intervals to generate a dynamic probability frequency band list.

7. A radio counter-jamming method for UAV security defense according to claim 6, characterized in that, The method of suppressing UAV communication signals and ensuring airspace security by using a radio countermeasure device, based on the probability value of the central frequency band in the airspace electromagnetic signal data belonging to the UAV communication frequency band, specifically includes: The radio countermeasure device is based on software-defined radio technology, which supports flexible switching of operating frequency bands within the 300MHz-6GHz frequency band range and can simultaneously transmit 1-4 sets of countermeasure signals in different frequency bands. Based on the data in the UAV communication frequency band library, pre-configure corresponding countermeasure signals for different frequency bands for each radio countermeasure device; The radio countermeasure device is equipped with an adaptive power amplifier, which can automatically adjust the transmission power of the countermeasure signal according to the distance and signal strength of the target drone; Based on the probability values ​​of the frequency bands in the airspace electromagnetic signal data belonging to the UAV communication frequency bands, the frequency band types with probability values ​​exceeding the preset value are filtered out. Based on the selected frequency band type, the corresponding radio countermeasure device is activated to actively transmit countermeasure signals to interfere with the drone's communication signals.

8. A radio counter-jamming device for drone security defense, used to implement the radio counter-jamming method for drone security defense as described in any one of claims 1-7, characterized in that, include: The library module is used to collect and record the commonly used communication frequency bands of existing UAVs and the signal characteristics of common frequency interference sources, and to construct a UAV communication frequency band library and an interference source feature library, respectively. The airspace monitoring module is used to deploy ultra-wideband radio frequency antenna devices in multiple directions, collect spectrum data in the airspace, transmit the data to the central processing center via a wired network, and establish an airspace electromagnetic signal dataset. The feature extraction module is used to construct an electromagnetic signal feature extraction model based on the spatial electromagnetic signal dataset, and to obtain the time-domain features, frequency hopping features and modulation mode features of the electromagnetic signal respectively. The interference signal removal module is used to construct an interference source identification model based on the interference source feature library and remove electromagnetic signals of interference sources from the spatial electromagnetic signal data. The countermeasure and feedback module is used to construct a UAV frequency band identification probability model based on a UAV communication frequency band library and a Bayesian algorithm to determine the probability that a frequency band in the airspace electromagnetic signal data belongs to the UAV communication frequency band; based on the probability value of the frequency band in the airspace electromagnetic signal data belonging to the UAV communication frequency band, it suppresses the UAV communication signal through a radio countermeasure device to ensure the safety of the airspace; based on the feedback of the airspace electromagnetic signal data, it comprehensively evaluates the radio countermeasure effect, and when the radio countermeasure fails, it forcibly triggers a strong magnetic excitation device and updates the database synchronously.

9. A radio counter-jamming device for UAV security defense according to claim 8, characterized in that, The countermeasure and feedback module specifically includes: The frequency band probability unit is used to construct a UAV frequency band identification probability model based on the UAV communication frequency band library and the Bayesian algorithm to determine the probability that the frequency band in the airspace electromagnetic signal data belongs to the UAV communication frequency band. A radio countermeasure unit is used to suppress UAV communication signals and ensure the safety of the airspace by using a radio countermeasure device based on the probability value that the central frequency band of the airspace electromagnetic signal data belongs to the UAV communication frequency band. The countermeasure effect feedback unit is used to comprehensively evaluate the radio countermeasure effect based on the feedback of spatial electromagnetic signal data, and to forcibly trigger the strong magnetic excitation device and update the database synchronously when the radio countermeasure fails.

Images