A method for cooperative control of detection and jamming for spread spectrum and frequency hopping unmanned aerial vehicles

Abstract

A kind of detection interference cooperative control method for spread spectrum, frequency hopping unmanned aerial vehicle, including real-time acquisition of the detection data of radar, form first identification information;Real-time acquisition of the detection data of photoelectric equipment, form second identification information;Real-time acquisition of the detection data of radio reconnaissance equipment deployed in multiple positions, obtain third classification identification result;The first identification information, second identification information and third classification identification result are input into data fusion system in real time, and dynamic comprehensive processing is carried out, obtains final identification information and generates interference radiation power regulation parameter, and dynamic generation control instruction is transmitted to interference equipment;Interference equipment releases interference information to target unmanned aerial vehicle, until realizing complete countermeasure to target unmanned aerial vehicle;The present application solves the defects of traditional means in timeliness, accuracy, systematization through the closed-loop architecture of "intelligent perception-collaborative decision-making-accurate execution", realizes the effective interference to frequency hopping, spread spectrum unmanned aerial vehicle, reduces collateral damage.

Description

Technical Field

[0001] This invention relates to the field of unmanned aerial vehicle (UAV) countermeasures technology, specifically to a detection and jamming cooperative control method for spread spectrum and frequency hopping UAVs. Background Technology

[0002] With the widespread adoption of drone technology, the security risks it poses are becoming increasingly severe. Radio detection and electronic jamming are effective means of countering drones. However, to enhance their anti-jamming capabilities, advanced drones extensively employ spread spectrum and frequency hopping communication technologies. These signals have low power spectral density and random carrier frequency hopping. In existing technologies, detection and jamming units are mostly independent, resulting in delayed system response, coarse jamming patterns (often omnidirectional and full-band jamming, causing significant collateral damage), and a lack of intelligent decision-making capabilities. This makes it impossible to achieve rapid, accurate, and adaptive responses, rendering traditional jamming ineffective against such drones.

[0003] For example, patent document CN120454917A discloses a method, system, and storage device for detecting and jamming unmanned aerial vehicles (UAVs). The UAV detection and jamming method includes the following steps: S1, detecting radio signals in real time using at least four pre-deployed receiving stations, and identifying the target UAV signal through spectrum analysis; S2, determining and continuously updating the spatial position of the target UAV based on the absolute time and frequency of the target UAV signal arriving at each of the receiving stations; S3, fitting the spatial position of the target UAV and predicting the target UAV path within a set time period; S4, determining the relative positional relationship between each countermeasure device and the target UAV, and determining the countermeasure device to be executed, the optimal position of the countermeasure device, the countermeasure azimuth angle, and the countermeasure area angle; S5, controlling the optimal countermeasure device to counter the target UAV.

[0004] For example, patent document CN115733575B discloses a drone detection and countermeasure system and method. This drone detection and countermeasure system includes a radio detection and jamming device, a display and control device, and a fixed jamming device. The radio detection and jamming device is used to detect drones and, in response to precision strike control commands issued by the display and control device, generates and transmits corresponding jamming signals. The display and control device, upon receiving drone information, determines the target drone and, based on the target drone's location information, generates precision strike control commands and sends them to the radio detection and jamming device, or generates directional jamming control commands and sends them to the fixed jamming device. The fixed jamming device, in response to directional jamming control commands issued by the display and control device, generates and transmits directional jamming signals. Therefore, there is an urgent need for a technical solution that can achieve intelligent linkage control that deeply integrates drone detection and jamming. Summary of the Invention

[0005] To address the technical problems of existing systems exhibiting slow response, coarse interference patterns, and a lack of intelligent decision-making capabilities, which prevent rapid, accurate, and adaptive countermeasures and thus hinder the effective countermeasures against UAVs employing spread spectrum and frequency hopping communication technologies, this invention provides a cooperative control method for detecting and interfering with spread spectrum and frequency hopping UAVs. The method includes: The radar detection data is acquired in real time, preprocessed, and the radar feature-level information of the target UAV is extracted and the first classification and identification result is obtained to form the first identification information; the first classification and identification result includes the target UAV's position, pitch, velocity, radar cross section and type information; the radar feature-level information includes micro-motion features, kinematic features and scattering features; The system acquires real-time detection data from optoelectronic devices, performs preprocessing, extracts optoelectronic feature-level information of the target UAV, and obtains a second classification and recognition result to form second recognition information. The second classification and recognition result includes the type and speed information of the target UAV. The optoelectronic feature-level information is a visual feature, including the UAV's shape, texture, motion, and thermal infrared. The system acquires real-time detection data from radio reconnaissance equipment deployed at multiple locations, preprocesses it, extracts reconnaissance feature-level information of the target UAV, performs classification and identification, obtains classification results from multiple locations, compares them with pre-established environmental and feature databases, obtains comparison results from multiple locations, performs fusion identification, and obtains a third classification identification result. The third classification identification result includes the target UAV's location information and communication parameters. The reconnaissance feature-level information includes transient features, error vector amplitude, carrier frequency offset, and phase noise. The communication parameters include communication frequency band and hopping rate. The first identification information, the second identification information, and the third classification identification result are input into the data fusion system in real time for dynamic comprehensive processing to obtain the final identification information of the target UAV and generate interference radiation power adjustment parameters; combined with the location information of the interference device, control commands are dynamically generated and transmitted to the interference device. The jamming device dynamically responds to the control command and releases jamming information to the target UAV until it completely counters the target UAV, thus completing the detection and jamming operation.

[0006] Furthermore, the process of obtaining the first classification and recognition result includes the following: The radar detection data is processed by removing the DC component, pulse compression, clutter and interference suppression, and constant false alarm rate (CFAR) processing to obtain the target UAV's trace. The points are correlated, constructed, filtered, and predicted to obtain the estimated motion parameters of the target UAV and form a flight path. The multidimensional features of the dots are extracted, combined with a convolutional neural network, and fused for recognition to classify the target UAV and generate a first classification result.

[0007] Furthermore, the process of forming the second identification information includes the following: The detection data from the optoelectronic device is subjected to denoising and non-uniformity correction image preprocessing and enhancement to obtain an enhanced image; Based on the enhanced images, moving targets are detected, and the visual features of the target drone are extracted; Based on the enhanced image, the YOLO algorithm is used to obtain the type and speed information of the target UAV, thereby enabling the type identification and tracking of the target UAV. The second identification information is formed based on the visual characteristics of the target drone and the type and speed information of the target drone.

[0008] Furthermore, in the process of obtaining classification results from multiple locations, classification and recognition are performed using a deep learning algorithm model; the deep learning algorithm model is pre-trained and learns from the environment. The pre-training involves training a deep learning model using existing UAV time-frequency map data and target data with similar communication systems. The environmental learning involves training by collecting the time-frequency characteristics, azimuth, and energy parameters of existing UAV abnormal signals.

[0009] Furthermore, the process of generating the interference radiation power adjustment parameters includes the following: Based on the communication parameters of the target UAV in the third classification identification result, a short-time Fourier transform (STFT) is performed to convert the communication parameters into a time-frequency diagram. The time-frequency graph is analyzed to obtain the target UAV's frequency band, hopping rate, and modulation scheme. Based on the target UAV's frequency band, hopping rate, and modulation method, determine the interference pattern and frequency band; Obtain the target UAV type and radial distance, combine them with the previously established power-type-distance mapping library, determine the radiated power, and generate interference radiated power adjustment parameters.

[0010] Furthermore, the parameter parsing is performed using clustering algorithms or deep learning algorithms. The clustering algorithms include the unsupervised clustering algorithm K-Means, the noisy spatial clustering algorithm DBSCAN, and the fuzzy C-means clustering algorithm FCM. The deep learning algorithms include convolutional neural networks, recurrent neural networks, and long short-term memory networks.

[0011] Furthermore, before achieving complete countermeasures against the target drone, the system continuously generates first identification information, second identification information, and third classification identification results to achieve continuous detection and tracking of the target drone. The data fusion system dynamically updates control commands to achieve precise interference with the target drone.

[0012] Furthermore, the complete countermeasure refers to the target UAV exhibiting behaviors such as returning to base, hovering, or crashing; the control commands include the target UAV's azimuth, pitch, jamming frequency band, jamming pattern, and radiated power.

[0013] Furthermore, the final identification information includes the target UAV's longitude, latitude, altitude, speed, type, frequency band, hop rate, and modulation method.

[0014] The beneficial effects of this invention are as follows: This invention acquires detection data from multiple sources, classifies and identifies target UAVs, and then uses a data fusion system to comprehensively process the identification information from the multiple sources to obtain the UAV's location and type information. Based on STFT+ clustering / deep learning algorithms, it analyzes the UAV's frequency band range and hopping rate, and combines this with a pre-established "power-type-distance" mapping library to generate interference radiation power adjustment parameters, achieving coordinated control with the jamming devices. Simultaneously, before achieving complete countermeasures, it acquires detection data from multiple sources in real time, enabling continuous detection and tracking of the target UAV. The comprehensive processing of the data fusion system is performed in real time, dynamically generating interference radiation power adjustment parameters to achieve precise jamming of the target UAV. Through a closed-loop architecture of "intelligent perception-collaborative decision-making-precise execution," it overcomes the shortcomings of traditional methods in terms of timeliness, accuracy, and systematization, achieving effective jamming of frequency-hopping and spread-spectrum UAVs and reducing collateral damage. Attached Figure Description

[0015] Figure 1 This is a flowchart of the detection and jamming cooperative control method for spread spectrum and frequency hopping UAVs provided by the present invention; Figure 2 This is a block diagram of the integrated processing logic of the data fusion system provided by the present invention. Detailed Implementation

[0016] The technical solution of the present invention is further described below, but the scope of protection is not limited to what is described.

[0017] This invention provides a detection and jamming cooperative control method for spread spectrum and frequency hopping UAVs, such as... Figure 1 As shown, the method includes: Step S100: Real-time acquisition of radar detection data, preprocessing, extraction of radar feature-level information of the target UAV, and acquisition of first classification and identification results to form first identification information; the first classification and identification results include the target UAV's position, pitch, velocity, radar cross section (RCS), and type information; the radar feature-level information includes micro-motion characteristics, kinematic characteristics, and scattering characteristics; the kinematic characteristics include radial velocity, acceleration, and trajectory; the micro-motion characteristics include micro-motion Doppler spectral lines, airframe vibration, and modulation periodic spectrum; the scattering characteristics include radar cross section and fluctuation characteristics; The process of obtaining the first classification and recognition result includes the following: The radar detection data is processed by removing the DC component, pulse compression, clutter and interference suppression, and constant false alarm rate (CFAR) processing to obtain the target UAV's trace. The points are correlated, established, filtered, and predicted to obtain the estimated motion parameters of the target UAV and form a flight path; the motion parameters of the target UAV include position, velocity, acceleration, and micro-doppler parameters; The multidimensional features of the dots are extracted and combined with convolutional neural networks or machine learning algorithms for fusion recognition to classify the target UAV and generate a first classification result. The multidimensional features of the dots include radial distance, azimuth angle, pitch angle, radial velocity, and echo amplitude. Step S200: Real-time acquisition of detection data from optoelectronic devices, preprocessing, extraction of optoelectronic feature-level information of the target UAV and acquisition of second classification recognition results to form second recognition information; the second classification recognition results include the type and speed information of the target UAV; the optoelectronic feature-level information is visual features, including UAV shape, texture, motion and thermal infrared; The process of forming the second identification information includes the following: The detection data from the optoelectronic device is subjected to denoising and non-uniformity correction image preprocessing and enhancement to obtain an enhanced image; Based on the enhanced images, moving targets are detected using inter-frame difference or optical flow methods, and the visual features of the target UAV are extracted. Based on enhanced images, YOLO or machine learning algorithms are used to obtain the type and speed information of the target drone, enabling type identification and tracking of the target drone; specifically: First, a convolutional neural network is used to extract features from the image, resulting in an S×S feature map. Then, the entire image is divided into S×S cells, with each cell responsible for detecting objects whose center falls within it, predicting the object's position and category information, thereby obtaining the target drone's type and speed information.

[0018] The second identification information is formed based on the visual characteristics of the target drone and the type and speed information of the target drone.

[0019] Step S300: Real-time acquisition of detection data from radio reconnaissance equipment deployed at multiple locations; preprocessing; extraction of reconnaissance feature-level information of the target UAV; classification and identification; acquisition of classification results from multiple locations; comparison with a pre-established environment database and feature database; acquisition of comparison results from multiple locations; fusion identification; obtaining a third classification identification result; the third classification identification result includes the target UAV's location information and communication parameters; the reconnaissance feature-level information includes transient features, error vector amplitude, carrier frequency offset, and phase noise; the communication parameters include communication frequency band and hopping rate; In the process of obtaining classification results from multiple locations, classification and recognition are performed using a deep learning algorithm model; the deep learning algorithm model is pre-trained and learns from the environment. The pre-training involves training a deep learning model using existing UAV time-frequency map data and target data with similar communication systems. The environmental learning involves training by collecting the time-frequency characteristics, azimuth, and energy parameters of existing UAV abnormal signals.

[0020] Step S400: The first identification information, the second identification information, and the third classification identification result are input into the data fusion system in real time for dynamic comprehensive processing to obtain the final identification information of the target UAV and generate interference radiation power adjustment parameters; combined with the location information of the interference device, control commands are dynamically generated and transmitted to the interference device; the final identification information includes the longitude, latitude, altitude, speed, type, frequency band, hopping rate, and modulation method of the target UAV; the control commands include the azimuth, pitch, interference frequency band, interference pattern, and radiation power of the target UAV; the logical framework of the comprehensive processing is as follows: Figure 2 As shown; The process of generating the interference radiation power adjustment parameters includes the following: Based on the communication parameters of the target UAV in the third classification identification result, a Short-Time Fourier Transform (STFT) is performed to convert the communication parameters into a time-frequency diagram; the STFT can be expressed as: (1) in, It is a signal The result after performing a Fourier transform This is a signal that needs to be processed. It is a shifted window function, which effectively suppresses signals outside the window function; The discrete form of STFT is represented as: (2) The time-frequency dataset after STFT conversion is defined as The objective function of the FCM algorithm is: (3) in, It is the number of sample categories. It is the total number of samples. It is a fuzzy factor. (k=1,2,3,…,c) are the cluster centers. Indicates the weighted membership degree. Data points To the cluster center The square of the Euclidean distance.

[0021] Membership degree The constraints are: (4) Due to the objective function Since it depends on both U and V and has no direct analytical solution, an alternating optimization method is used to solve it. The membership degree iterative update expression is: (5) The iterative update expression for cluster centers is: (6) After the iteration converges, all data points are determined according to the membership matrix. Divide into c disjoint subsets: (7) in It contains all the assigned first A drone data point, from Extracting frequency values Composition of frequency set Calculate respectively The maximum and minimum frequencies are used to determine the frequency band range of the frequency-hopping UAV. Similarly from The timestamps can be extracted to calculate the jump speed.

[0022] The formula for calculating the electromagnetic susceptibility field strength is shown below: (8) In the formula, E is the threshold of electromagnetic susceptibility field strength, in V / m; Pin — Injected transmit antenna power, W; Gt — Transmit antenna gain; r—electromagnetic sensitivity distance, m.

[0023] The effective interference power against UAVs can be calculated in reverse using the formula for electromagnetic sensitive field strength. The time-frequency graph is analyzed to obtain the frequency band, hopping speed, and modulation mode of the target UAV. The parameter analysis is performed using a clustering algorithm or a deep learning algorithm. The clustering algorithm includes the unsupervised clustering algorithm K-Means, the noisy spatial clustering algorithm DBSCAN, and the fuzzy C-means clustering algorithm FCM. The deep learning algorithm includes convolutional neural networks, recurrent neural networks, and long short-term memory networks.

[0024] Based on the target UAV's frequency band, hopping rate, and modulation method, determine the interference pattern and frequency band; The target UAV's type and radial distance are obtained. Combined with a pre-established power-type-distance mapping library, the radiated power is determined, and interference radiated power adjustment parameters are generated. This achieves effective interference with the UAV while avoiding excessive power from affecting other equipment.

[0025] The power-type-distance mapping library is established based on the sensitivity thresholds of various types of UAVs obtained from previous experiments; In step S500, the jamming device dynamically responds to the control command and releases jamming information to the target UAV until the target UAV is completely countered, thus completing this detection and jamming operation; the complete countermeasure refers to the target UAV exhibiting behaviors such as returning to home, hovering, or crashing.

[0026] Before achieving complete countermeasures against the target drone, the system continuously generates first identification information, second identification information, and third classification identification results to achieve continuous detection and tracking of the target drone. The data fusion system dynamically updates control commands to achieve precise interference with the target drone.

[0027] The above-disclosed embodiments are merely specific examples of the present invention. However, the present invention is not limited thereto, and any variations that can be conceived by those skilled in the art should fall within the protection scope of the present invention.

Claims

1. A detection and jamming cooperative control method for spread spectrum and frequency hopping UAVs, characterized in that, The method includes: The radar detection data is acquired in real time, preprocessed, and the radar feature-level information of the target UAV is extracted and the first classification and identification result is obtained to form the first identification information; the first classification and identification result includes the target UAV's position, pitch, velocity, radar cross section and type information; the radar feature-level information includes micro-motion features, kinematic features and scattering features; The system acquires real-time detection data from optoelectronic devices, performs preprocessing, extracts optoelectronic feature-level information of the target UAV, and obtains a second classification and recognition result to form second recognition information. The second classification and recognition result includes the type and speed information of the target UAV. The optoelectronic feature-level information is a visual feature, including the UAV's shape, texture, motion, and thermal infrared. The system acquires real-time detection data from radio reconnaissance equipment deployed at multiple locations, preprocesses it, extracts reconnaissance feature-level information of the target UAV, performs classification and identification, obtains classification results from multiple locations, compares them with pre-established environmental and feature databases, obtains comparison results from multiple locations, performs fusion identification, and obtains a third classification identification result. The third classification identification result includes the target UAV's location information and communication parameters. The reconnaissance feature-level information includes transient features, error vector amplitude, carrier frequency offset, and phase noise. The communication parameters include communication frequency band and hopping rate. The first identification information, the second identification information, and the third classification identification result are input into the data fusion system in real time for dynamic comprehensive processing to obtain the final identification information of the target UAV and generate interference radiation power adjustment parameters; combined with the location information of the interference device, control commands are dynamically generated and transmitted to the interference device. The jamming device dynamically responds to the control command and releases jamming information to the target UAV until it completely counters the target UAV, thus completing the detection and jamming operation.

2. The detection and jamming cooperative control method for spread spectrum and frequency hopping UAVs as described in claim 1, characterized in that, The process of obtaining the first classification and recognition result includes the following: The radar detection data is processed by removing the DC component, pulse compression, clutter and interference suppression, and constant false alarm rate (CFAR) processing to obtain the target UAV's trace. The points are correlated, constructed, filtered, and predicted to obtain the estimated motion parameters of the target UAV and form a flight path. The multidimensional features of the dots are extracted, combined with a convolutional neural network, and fused for recognition to classify the target UAV and generate a first classification result.

3. The detection and jamming cooperative control method for spread spectrum and frequency hopping UAVs as described in claim 1, characterized in that, The process of forming the second identification information includes the following: The detection data from the optoelectronic device is subjected to denoising and non-uniformity correction image preprocessing and enhancement to obtain an enhanced image; Based on the enhanced images, moving targets are detected, and the visual features of the target drone are extracted; Based on the enhanced image, the YOLO algorithm is used to obtain the type and speed information of the target UAV, thereby enabling the type identification and tracking of the target UAV. The second identification information is formed based on the visual characteristics of the target drone and the type and speed information of the target drone.

4. The detection and jamming cooperative control method for spread spectrum and frequency hopping UAVs as described in claim 1, characterized in that, In the process of obtaining classification results from multiple locations, classification and recognition are performed using a deep learning algorithm model; the deep learning algorithm model is pre-trained and learns from the environment. The pre-training involves training a deep learning model using existing UAV time-frequency map data and target data with similar communication systems. The environmental learning involves training by collecting the time-frequency characteristics, azimuth, and energy parameters of existing UAV abnormal signals.

5. The detection and jamming cooperative control method for spread spectrum and frequency hopping UAVs as described in claim 1, characterized in that, The process of generating the interference radiation power adjustment parameters includes the following: Based on the communication parameters of the target UAV in the third classification identification result, a short-time Fourier transform (STFT) is performed to convert the communication parameters into a time-frequency diagram. The time-frequency graph is analyzed to obtain the target UAV's frequency band, hopping rate, and modulation scheme. Based on the target UAV's frequency band, hopping rate, and modulation method, determine the interference pattern and frequency band; Obtain the target UAV type and radial distance, combine them with the previously established power-type-distance mapping library, determine the radiated power, and generate interference radiated power adjustment parameters.

6. The detection and jamming cooperative control method for spread spectrum and frequency hopping UAVs as described in claim 5, characterized in that, The parameter analysis is performed using clustering algorithms or deep learning algorithms. The clustering algorithms include unsupervised clustering algorithm K-Means, noisy spatial clustering algorithm DBSCAN, and fuzzy C-means clustering algorithm FCM. The deep learning algorithms include convolutional neural networks, recurrent neural networks, and long short-term memory networks.

7. The detection and jamming cooperative control method for spread spectrum and frequency hopping UAVs as described in claim 1, characterized in that, Before achieving complete countermeasures against the target drone, the system continuously generates first identification information, second identification information, and third classification identification results to achieve continuous detection and tracking of the target drone. The data fusion system dynamically updates control commands to achieve precise interference with the target drone.

8. The detection and jamming cooperative control method for spread spectrum and frequency hopping UAVs as described in claim 7, characterized in that, The complete countermeasure refers to the target UAV returning to base, hovering, or crashing; the control commands include the target UAV's azimuth, pitch, jamming frequency band, jamming pattern, and radiated power.

9. The detection and jamming cooperative control method for spread spectrum and frequency hopping UAVs as described in claim 1, characterized in that, The final identification information includes the target UAV's longitude, latitude, altitude, speed, type, frequency band, hop rate, and modulation method.

Images