Unmanned aerial vehicle detection method and system, electronic equipment and storage medium

Abstract

The invention relates to the technical field of unmanned aerial vehicle detection, in particular to an unmanned aerial vehicle detection method and system, electronic equipment and a storage medium, and the method comprises the steps: obtaining a to-be-detected radio frequency signal; the radio frequency signal is obtained by filtering an obtained original radio frequency signal based on a filter bank; generating a time-frequency image based on the radio frequency signal; detecting the time-frequency image by using a preset detection model, and generating a detection bounding box of the radio frequency signal on the time-frequency image; correcting the time domain component and the frequency domain component of the detection bounding box based on the signal characteristics of the radio frequency signal to obtain a corrected detection bounding box; and based on the corrected detection bounding box, selecting an image area from the time-frequency image, and carrying out unmanned aerial vehicle detection. The method is suitable for the unmanned aerial vehicle detection process and is used for improving the unmanned aerial vehicle detection accuracy.

Description

Technical Field

[0001] This invention relates to the field of unmanned aerial vehicle (UAV) detection technology, specifically to a UAV detection method, system, electronic device, and storage medium. Background Technology

[0002] With the rapid development of the low-altitude economy, drones are widely used in many fields, but they also bring numerous security threats, such as unauthorized malicious flights and eavesdropping attacks. Current drone detection methods mainly include the following: (1) Radar-based UAV detection method: UAVs are detected by using radar echo signals, and the presence, type and motion status of the UAV are determined by analyzing the characteristics of the echo. However, due to their extremely small radar cross-section, micro UAVs may not be effectively detected in some cases, leading to missed detections. For example, some micro UAVs weighing less than 250 grams have very weak radar echo signals, which are difficult to distinguish from background noise.

[0003] (2) Vision-based UAV detection methods: Deep learning models, such as You Only Look Once (YOLO) and faster region-based convolutional neural networks, are used to analyze images and videos to identify UAVs. Under ideal conditions, these algorithms can achieve good detection results. However, in practical applications, they face many challenges. Different weather conditions, such as rain, fog, and snow, can seriously affect image quality and reduce the detection accuracy of the algorithm. Differences in image resolution can also affect the detection results; low-resolution images may not provide enough detail to accurately identify UAVs. In addition, the flight distance of the UAV and occlusion in the scene can also interfere with the normal operation of the algorithm. When the UAV is far away or is occluded by objects such as buildings and trees, the detection difficulty will increase significantly.

[0004] (3) Acoustic-based UAV detection method: This method identifies UAVs by capturing the sound generated by their propellers using a microphone array. However, this method is highly sensitive to environmental noise. In noisy environments, such as city streets or near factories, the ambient noise may mask the UAV's sound signal, making accurate identification impossible. Flight distance also affects the detection effect; as the UAV's flight distance increases, the sound signal gradually weakens, increasing the difficulty of detection. In addition, some UAVs employ noise reduction technology, further reducing the detectability of their sound signals. Summary of the Invention

[0005] To address the problem of potential deviations in drone detection results in existing technologies, this invention provides a drone detection method, system, electronic device, and storage medium.

[0006] This invention is achieved through the following technical solution: A method for detecting unmanned aerial vehicles (UAVs), comprising: Acquire the radio frequency signal to be detected; Based on the radio frequency signal, a time-frequency image is generated; The time-frequency image is detected using a preset detection model, and a detection bounding box for the radio frequency signal is generated on the time-frequency image. Based on the signal characteristics of the radio frequency signal, the time-domain and frequency-domain components of the detection bounding box are corrected to obtain the corrected detection bounding box. UAV detection is performed on the image regions selected from the time-frequency images based on the corrected detection bounding boxes.

[0007] Preferably, the step of correcting the time-domain and frequency-domain components of the detection bounding box based on the signal characteristics of the radio frequency signal includes: Based on the estimated bandwidth of the radio frequency signal, the frequency domain components of the detection bounding box are corrected. The temporal components of the detection bounding box are corrected based on the Zadov-Zhu ZC sequence of the radio frequency signal.

[0008] Preferably, the step of correcting the frequency domain components of the detection bounding box based on the estimated bandwidth of the radio frequency signal includes: To obtain the estimated bandwidth, including: The radio frequency signal is divided into lengths of Window function Divided into Overlapping sequences: ; in, Indicates the first division Overlapping sequences at position The value at; Indicates that the radio frequency signal is in The value at; Indices representing the overlapping sequences; Indicates the offset. ; Based on the above The power spectrum of each overlapping sequence is calculated using the following formula: ; in, This represents the power spectrum of the radio frequency signal; Indicates angular frequency; Based on the power spectrum, the initial estimated bandwidth is determined according to the following formula: ; in, This indicates the initial estimated bandwidth; express The maximum frequency corresponding to the maximum power; express The minimum frequency corresponding to the maximum power; Based on the initial estimated bandwidth, the estimated bandwidth is calculated according to the following formula: ; in, Indicates the total number of subcarriers; Indicates the number of subcarriers used for data transmission; If the frequency band Fd of the detected radio frequency signal is not within the specified frequency band of the UAV communication protocol. Within this range, the frequency domain components of the detected bounding box are corrected according to the following formula: ; ; in, Indicates the first A specified frequency band, , express The center frequency; This represents the maximum frequency in the frequency domain component; This represents the minimum frequency in the frequency domain component; This indicates the estimated bandwidth.

[0009] Preferably, the correction of the temporal components of the detection bounding box based on the Zadov-Zhu ZC sequence of the radio frequency signal includes: The ZC sequence of the radio frequency signal is determined according to the following formula: ; in, This represents the ZC sequence; Indicates the root index; Indicates the sequence length of the ZC sequence; The ZC sequence and the radio frequency signal are cross-correlated according to the following formula: ; in, This represents the cross-correlation value between the ZC sequence and the radio frequency signal; Indicates the length of the window; The total length of the radio frequency signal; Indicates that the radio frequency signal is in The value; Based on the cross-correlation value Peak index The temporal components of the detected bounding box are corrected according to the following formula: ; in, This represents the minimum time in the time-domain component; Indicates the preset coefficient; This indicates the length of the cyclic prefix (CP) under normal circumstances. Indicates the length of the extended loop prefix; This represents the maximum time in the time-domain component; Indicates the specified duration of a broadcast frame. , This indicates the number of symbols in a broadcast frame. Indicates the CP length; This represents the total number of subcarriers of the radio frequency signal; This indicates the bandwidth of the radio frequency signal.

[0010] Preferably, the correction of the temporal components of the detection bounding box based on the Zadov-Zhu ZC sequence of the radio frequency signal includes: The ZC sequence of the radio frequency signal is determined according to the following formula: ; in, This represents the ZC sequence; Indicates the root index; Indicates the sequence length of the ZC sequence; The ZC sequence and the radio frequency signal are cross-correlated according to the following formula: ; in, This represents the cross-correlation value between the ZC sequence and the radio frequency signal; Indicates the length of the window; The total length of the radio frequency signal; Indicates that the radio frequency signal is in The value; Based on the cross-correlation value Peak index The temporal components of the detected bounding box are corrected according to the following formula: ; in, This represents the minimum time in the time-domain component; Indicates the preset coefficient; This indicates the length of the cyclic prefix (CP) under normal circumstances. Indicates the length of the extended loop prefix; This represents the maximum time in the time-domain component; Indicates the specified duration of a broadcast frame. , This indicates the number of symbols in a broadcast frame. Indicates the CP length; This represents the total number of subcarriers of the radio frequency signal; This indicates the bandwidth of the radio frequency signal.

[0011] Preferably, for cross-correlation values Make corrections, including: cross-correlation value Divide into Q non-overlapping sequences, each of length P, and PQ = LV; Calculate the energy value of each sequence using the following formula: ; in, Indicates energy value; Indicates a sequence index; The interference signal index set is determined according to the following formula. : ; in, This indicates the preset weighting factor. This represents the average energy value of all sequences. right Correct according to the following formula: ; in, This represents the corrected cross-correlation value; This represents the cross-correlation value before correction.

[0012] Preferably, the step of performing UAV detection based on the image region selected from the time-frequency image using the corrected detection bounding box includes: Based on the radio frequency signal corresponding to the image region, a detection radio frequency signal is obtained; The frequency of the detected radio frequency signal is shifted from the carrier frequency to near zero frequency to obtain the initial baseband signal; After filtering the initial baseband signal using a low-pass filter, the initial baseband signal is resampled according to a preset sampling rate to obtain the target baseband signal. ; The autocorrelation value of the target baseband signal over the CP length range is calculated using the following formula: ; in, This represents the autocorrelation value of the target baseband signal within the CP length range; Indicates the CP length; This indicates the length of an OFDM symbol excluding the CP portion; based on The peak index is used to determine the starting index of the OFDM symbol, and the complete OFDM symbol is truncated based on the starting index; the OFDM symbol includes a CP part and a main body part; The time-domain signal of the main body is extracted and transformed into the frequency domain through a fast Fourier transform to obtain the frequency domain signal; The phase value of the pilot subcarrier is extracted from the frequency domain signal, and the phase difference between the current symbol and the same pilot subcarrier in the previous symbol is calculated. ; The frequency shift estimate is calculated using the following formula: ; in, This represents the estimated frequency offset. Indicates the number of points in the Fast Fourier Transform; Indicates the duration of an OFDM symbol; Based on the frequency offset estimate, the target baseband signal is phase-corrected to obtain the corrected target baseband signal; The corrected target baseband signal is used as a bitstream and descrambled based on the Gold sequence to restore the original bitstream. The original bitstream is corrected using a Turbo decoder to recover the payload information; UAV detection is performed based on the payload information.

[0013] An electronic device includes: a processor and a memory; The memory stores instructions that the processor can execute; When the processor is configured to execute the instructions, it causes the electronic device to implement the method.

[0014] A readable storage medium comprising: software instructions; When the software instructions are executed in an electronic device, the electronic device causes the electronic device to implement the method described.

[0015] A computer program product, comprising: computer instructions; When the computer instructions are executed in an electronic device, the electronic device causes the electronic device to perform the method described.

[0016] Compared with the prior art, the present invention has the following beneficial effects: This invention discloses a method for detecting unmanned aerial vehicles (UAVs). The method involves acquiring a radio frequency (RF) signal to be detected. The RF signal is obtained by filtering the acquired raw RF signal using a filter bank. The filter bank includes at least one filter. The at least one filter is used to filter out interference signals from the raw RF signal that differ from the frequency bands and signal bandwidths specified in the UAV communication protocol. Based on the RF signal, a time-frequency image is generated. A preset detection model is used to detect the time-frequency image, generating a detection bounding box for the RF signal on the time-frequency image. Based on the signal characteristics of the RF signal, the time-domain and frequency-domain components of the detection bounding box are corrected to obtain a corrected detection bounding box. Based on the corrected detection bounding box, an image region selected from the time-frequency image is used for UAV detection. This method can use a preset detection model to detect the time-frequency image, generating a detection bounding box for the RF signal on the time-frequency image; and correct the time-domain and frequency-domain components of the detection bounding box based on the signal characteristics of the RF signal to obtain a corrected detection bounding box, thus providing a method for UAV detection based on radio frequency signals.

[0017] Furthermore, existing methods for correcting the pixel set of the detection bounding box (such as dehazing or noise reduction algorithms) in related technologies cannot reliably and consistently improve the detection accuracy of broadcast frames. This is because in time-frequency images, the characteristics of interference signals and broadcast frames are extremely similar, making pixel-level image reconstruction operations difficult to effectively support semantic-level target detection. In complex scenarios with low signal-to-noise ratios and overlapping signals, such reconstruction algorithms not only fail to improve detection accuracy but also reduce the detection accuracy of frequency-hopping spread spectrum (FHSS) signals and broadcast frames. This invention can correct the detection bounding box based on the signal characteristics of the radio frequency signal, achieving data-level decoupling in the detection of weak targets, which can significantly improve the detection accuracy of the model.

[0018] Furthermore, the cross-correlation value is corrected because sequences with high energy accumulation may correspond to interference signals. By using the above formula, sequences with high energy accumulation can be cleared to zero, thereby eliminating interference signals. Sequences with lower energy and more normal characteristics are retained to correct the time-domain components, which can improve the accuracy of correcting the time-domain components.

[0019] Furthermore, the target baseband signal is phase-corrected based on the frequency offset estimate to obtain the corrected target baseband signal; the corrected target baseband signal is then descrambled using a Gold sequence as a bitstream to restore the original bitstream; the original bitstream is then error-corrected using a Turbo decoder to recover the payload information; and the UAV is then detected based on the payload information. Attached Figure Description

[0020] Figure 1 This is a flowchart illustrating the drone detection method provided in an embodiment of the present invention; Figure 2 This is a schematic diagram of subcarrier distribution provided in an embodiment of the present invention; Figure 3 This is a schematic diagram illustrating the relationship between the estimation error of the estimated bandwidth and the signal-to-noise ratio when the sampling frequency is 100MHz, as provided in an embodiment of the present invention. Figure 4 This is a schematic diagram illustrating the relationship between the estimation error of the total number of subcarriers and the signal-to-noise ratio when the sampling frequency is 100MHz, as provided in an embodiment of the present invention. Figure 5 This is a schematic diagram of the composition of the UAV detection device provided in an embodiment of the present invention; Figure 6 This is a schematic diagram of the composition of an electronic device provided in an embodiment of the present invention. Detailed Implementation

[0021] The present invention will be further described in detail below with reference to specific embodiments. These descriptions are for illustrative purposes only and not for limiting the invention. Hereinafter, the terms "first," "second," and "third," etc., are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Therefore, a feature defined with "first," "second," or "third," etc., may explicitly or implicitly include one or more of that feature.

[0022] With the rapid development of the low-altitude economy, drones are widely used in many fields, but they also bring many security threats, such as unauthorized malicious flights and eavesdropping attacks. The drone detection method provided in this embodiment of the invention uses a drone detection device as the executing entity. This drone detection device can be an electronic device with computing processing capabilities, such as a computer or server. Optionally, the drone detection device can also be a processor (e.g., a central processing unit, CPU) in the aforementioned electronic device; or, the drone detection device can also be a software system or platform deployed in the aforementioned electronic device; or, the drone detection device can also be an application (APP) with drone detection functionality installed in the aforementioned electronic device; or, the drone detection device can also be a functional module in the aforementioned electronic device used to execute the drone detection method, etc. This embodiment of the invention does not impose any limitations on these aspects.

[0023] This invention discloses a method for detecting unmanned aerial vehicles (UAVs), referring to... Figure 1 S101, Acquire the radio frequency signal to be detected.

[0024] The radio frequency signal is obtained by filtering the acquired raw radio frequency signal based on a filter bank. The filter bank includes at least one filter, which is used to filter out interference signals from the raw radio frequency signal that have different frequency bands and signal bandwidths specified in the UAV communication protocol.

[0025] In some embodiments, before filtering the original radio frequency signal, the UAV detection device may also perform a short-time Fourier transform calculation according to the following formula: Formula (1) In formula (1), This represents the result of the short-time Fourier transform calculation. This represents the original radio frequency signal. This represents the window function at time t. Indicates the length of the window. This represents the frequency parameter.

[0026] S102. Generate a time-frequency image based on the radio frequency signal.

[0027] In the time-frequency image, the horizontal axis represents time (in milliseconds ms), and the vertical axis represents frequency (in megahertz MHz).

[0028] S103. Use a preset detection model to detect the time-frequency image and generate a detection bounding box for the radio frequency signal on the time-frequency image.

[0029] Optionally, the preset detection model can be a YOLOv7 model, a YOLOv5 model, a Faster R-CNN model, or an EfficientDet model, etc. This embodiment of the invention does not limit the specific type of the preset detection model.

[0030] S104. Based on the signal characteristics of the radio frequency signal, the time-domain and frequency-domain components of the detection bounding box are corrected to obtain the corrected detection bounding box.

[0031] In some embodiments, the UAV detection device can correct the time-domain and frequency-domain components of the detected bounding box separately. In this case, S104 may specifically include the following steps: Step 1: Correct the frequency domain components of the detection bounding box based on the estimated bandwidth of the radio frequency signal.

[0032] As an example, if the frequency band Fd of the detected radio frequency signal is not within the specified frequency band of the UAV communication protocol... Within this range, the UAV detection device can correct the frequency domain components of the detection bounding box according to the following formula: Formula (2) Formula (3) In formulas (2) and (3), Indicates the first A specified frequency band, , express The center frequency; This represents the maximum frequency in the frequency domain component; Represents the minimum frequency in the frequency domain component; This indicates the estimated bandwidth.

[0033] In some embodiments, before performing corrections according to formulas (2) and (3) above, the UAV detection device may also calculate the estimated bandwidth according to the following process. : The radio frequency signal is divided into lengths of window function Divided into Overlapping sequences: Formula (4) In formula (4), Indicates the first division Overlapping sequences at position The value at; Indicates radio frequency signal at The value at; Indices representing the overlapping sequences; Indicates the offset. .

[0034] based on The power spectrum of each overlapping sequence is calculated using the following formula: Formula (5) In formula (5), Represents the power spectrum of a radio frequency signal; It represents angular frequency.

[0035] The initial estimated bandwidth is determined based on the power spectrum using the following formula: Formula (6) In formula (6), Indicates the initial estimated bandwidth; express The maximum frequency corresponding to the maximum power; express The minimum frequency corresponding to the maximum power; Based on the initial estimated bandwidth, the bandwidth is estimated using the following formula: Formula (7) In formula (7), Indicates the total number of subcarriers; This indicates the number of subcarriers used for data transmission.

[0036] For example, Figure 2 This is a schematic diagram of subcarrier distribution provided in an embodiment of the present invention, such as... Figure 2 As shown, a symbol can include both virtual subcarriers and subcarriers used for transmitting data. This distribution characteristic has a significant impact on the calculation of the actual signal bandwidth, serving as a key reference for accurately estimating the signal bandwidth and providing important information for modulation parameter estimation.

[0037] For example, Figure 3 This diagram illustrates the relationship between the estimation error of the estimated bandwidth and the signal-to-noise ratio when the sampling frequency is 100MHz, as provided in an embodiment of the present invention. Figure 3 As shown, Figure 3 The figure shows the relationship between the estimation error and signal-to-noise ratio for the estimated bandwidth with different total subcarrier numbers. As can be seen from the figure, the estimation error does not exceed 120 kHz when using commonly used subcarrier numbers N=1024 and N=2048, providing a quantitative reference for the accuracy of signal bandwidth estimation.

[0038] In some embodiments, the UAV detection device can also calculate the total number of subcarriers of the radio frequency signal according to the following formula: The radio frequency signal is autocorrelation calculated using the following formula: Formula (8) In formula (8), Indicates the autocorrelation value; Indicates the length of the window; Indicates radio frequency signal; Indicates a time index; Indicates signal The conjugate of complex numbers.

[0039] Determine the autocorrelation value peak position .

[0040] Among them, when and hour, Will A peak value appears at [location]. Simultaneously, the estimated signal bandwidth [is also mentioned]. The number of subcarriers N satisfies They can be mutually verified. Indicates the length of the symbol.

[0041] Based on peak position Calculate the total number of subcarriers using the following formula: Formula (9) In formula (9), Indicates the sampling frequency; This indicates the bandwidth of the radio frequency signal.

[0042] For example, Figure 4 This diagram illustrates the relationship between the estimation error of the total number of subcarriers and the signal-to-noise ratio when the sampling frequency is 100MHz, as provided in an embodiment of the present invention. Figure 4 As shown, Figure 4 The figure shows the relationship between the error of the total number of carriers estimation and the signal-to-noise ratio (SNR). Although there is a certain estimation error at low SNR, the estimation becomes more accurate as channel conditions improve. Since it is usually taken as a power of 2, even with errors at low SNR, it can be accurately estimated within a finite range, which provides a reference for estimating the number of subcarriers.

[0043] Step 2: Correct the temporal components of the detected bounding box based on the Zadoff-Chu (ZC) sequence of the radio frequency signal.

[0044] As an example, the process of correcting the time-domain components by a drone detection device may include: The ZC sequence of the radio frequency signal is determined according to the following formula: Formula (10) In formula (10), Represents a ZC sequence; Indicates the root index; This indicates the sequence length of the ZC sequence.

[0045] The ZC sequence and the radio frequency signal are cross-correlated according to the following formula: Formula (11) In formula (11), Indicates the cross-correlation value; Indicates the length of the window; The total length of the radio frequency signal; Indicates radio frequency signal at The value of .

[0046] Based on cross-correlation value Peak index The temporal components of the detected bounding box are corrected according to the following formula: Formula (12) In formula (12), Represents the minimum time in the time-domain component; Indicates the preset coefficient; This indicates the length of the cyclic prefix (CP) under normal circumstances. Indicates the length of the extended loop prefix; This represents the maximum time in the time-domain component; Indicates the specified duration of a broadcast frame. , This indicates the number of symbols in a broadcast frame. Indicates the CP length; Indicates the total number of subcarriers in the radio frequency signal; This indicates the bandwidth of the radio frequency signal.

[0047] In some embodiments, before correcting the temporal components of the detection bounding box, the UAV detection device may also correct the cross-correlation value according to the following procedure: cross-correlation value Divide into Q non-overlapping sequences, each with length P, and PQ = LV.

[0048] Calculate the energy value of each sequence using the following formula: Formula (13) in, Indicates energy value; Indicates a sequence index.

[0049] The interference signal index set is determined according to the following formula. : Formula (14) in, This indicates the preset weighting factor. This represents the average energy value of all sequences. right Correct according to the following formula: Formula (15) in, This represents the corrected cross-correlation value; This represents the cross-correlation value before correction.

[0050] It should be understood that sequences with high energy accumulation may correspond to interference signals. By combining the above formulas (13) to (15), sequences with high energy accumulation can be cleared to zero, thereby eliminating interference signals. Sequences with lower energy and more normal characteristics can be retained to correct time-domain components, which can improve the accuracy of correcting time-domain components.

[0051] S105. Based on the image region selected from the time-frequency image after correction of the detection bounding box, perform UAV detection.

[0052] As an example, a drone detection device can extract features of the drone's radio frequency signal (or broadcast frame signal) from a time-frequency image based on a corrected detection bounding box. These features include specific frequency patterns, time-domain waveform characteristics, and signal energy distribution. For instance, it can extract characteristic parameters such as pulse shape and period in the time domain, and center frequency and bandwidth in the frequency domain. The extracted features are then compared with pre-stored drone broadcast frame feature templates. These templates can be standard features of different types of drone broadcast frames. By calculating similarity (e.g., using correlation coefficients, Euclidean distance, etc.), it is determined whether the detected signal matches known drone broadcast frame features. If the similarity exceeds a set threshold, a drone signal is considered detected.

[0053] As an example, the above S105 may specifically include the following steps: S1. Based on the radio frequency signal corresponding to the image region, obtain the detection radio frequency signal.

[0054] S2. Shift the frequency of the detected radio frequency signal from the carrier frequency to near zero frequency to obtain the initial baseband signal.

[0055] S3. After filtering the initial baseband signal with a low-pass filter, the initial baseband signal is resampled according to a preset sampling rate to obtain the target baseband signal. ; S4. Calculate the autocorrelation value of the target baseband signal within the CP length range using the following formula: Formula (16) In formula (16), This represents the autocorrelation value of the target baseband signal within the CP length range; Indicates the CP length; This indicates the length of the Orthogonal Frequency Division Multiplexing (OFDM) symbol excluding the CP portion. This indicates the first time index, used to control the starting position for calculating the signal. This represents the second time index, with a value range of 1 to... It is used to perform summation operations on the target sent signals within a limited CP length range.

[0056] S5, based on The peak index is used to determine the starting index of the OFDM symbol, and the complete OFDM symbol is truncated based on the starting index.

[0057] The OFDM symbol includes a CP part and a main body part.

[0058] S6. Extract the time-domain signal of the main part, and convert it to the frequency domain through a fast Fourier transform to obtain the frequency domain signal.

[0059] S7. Extract the phase value of the pilot subcarrier from the frequency domain signal, and calculate the phase difference between the current symbol and the same pilot subcarrier in the previous symbol. .

[0060] S8. Calculate the frequency offset estimate using the following formula: Formula (17) In formula (17), This represents the estimated frequency offset. Indicates the number of points in the Fast Fourier Transform; Indicates the duration of an OFDM symbol.

[0061] S9. Perform phase correction on the target baseband signal based on the frequency offset estimate to obtain the corrected target baseband signal.

[0062] As an example, drone detection devices can be based on frequency offset estimates. Generate compensation factor The target baseband signal is obtained by comparing the compensation factor with the phase difference of the target baseband signal.

[0063] S10. Descramble the corrected target baseband signal as a bitstream based on the Gold sequence to restore the original bitstream.

[0064] S11. The original bitstream is corrected using the Turbo decoder to recover the payload information.

[0065] S12. Detect UAVs based on payload information.

[0066] As an example, payload information can be specifically shown in Table 1 below: Table 1

[0067] In the unmanned aerial vehicle (UAV) detection method provided in this embodiment of the invention, the UAV detection device can use a preset detection model to detect the time-frequency image and generate a detection bounding box of the radio frequency signal on the time-frequency image; based on the signal characteristics of the radio frequency signal, the time-domain component and frequency-domain component of the detection bounding box are corrected to obtain the corrected detection bounding box, thereby providing a UAV detection method based on radio frequency signals.

[0068] Furthermore, existing methods for correcting the pixel set of the detection bounding box (such as dehazing or noise reduction algorithms) in related technologies cannot reliably and consistently improve the detection accuracy of broadcast frames. This is because in time-frequency images, the characteristics of interference signals and broadcast frames are extremely similar, making pixel-level image reconstruction operations difficult to effectively support semantic-level target detection. In complex scenarios with low signal-to-noise ratios and overlapping signals, such reconstruction algorithms not only fail to improve detection accuracy but also reduce the detection accuracy of frequency-hopping spread spectrum (FHSS) signals and broadcast frames. This invention can correct the detection bounding box based on the signal characteristics of the radio frequency signal, achieving data-level decoupling in the detection of weak targets, which can significantly improve the detection accuracy of the model.

[0069] In some embodiments of the present invention, the sampling duration of the radio frequency (RF) signal can be dynamically adjusted according to different detection requirements. After RF signals with different sampling durations are converted into time-frequency images, the presentation of the broadcast frame in the time-frequency image differs, leading to variations in detection accuracy and speed. When the sampling duration is 1ms, the broadcast frame occupies a large area in the time-frequency image. At this time, only weak target detection under low signal-to-noise ratio needs to be considered, resulting in low time-domain parameter correction costs and the highest FPS. However, it can only handle RF signals with a duration of 55ms, leading to high detection latency. As the sampling duration increases, both detection accuracy and FPS decrease. At a sampling duration of 50ms, the FPS drops to its lowest point. Although the high cost of time-domain parameter correction for processing small and weak targets prevents it from meeting the 20FPS detection requirement, the detection latency of only 100ms per second is still acceptable. This invention establishes a trade-off between detection accuracy and speed by detecting broadcast frames from RF signals with different sampling durations, meeting the detection needs of different UAV monitoring scenarios.

[0070] As an example, in close-range, high-dynamic scenarios such as drone competition areas and near-airport protection zones, a sampling duration of 1ms can be used. In mid-range, routine monitoring scenarios in areas such as urban parks and scenic spots, a sampling duration of 5-10ms can be used. In long-range, low-dynamic scenarios, a sampling duration of 50ms can be used.

[0071] The foregoing mainly describes the solutions provided by the embodiments of the present invention from a methodological perspective. To achieve the above functions, the UAV detection module includes hardware structures and / or software modules corresponding to the execution of each function. Those skilled in the art should readily recognize that, in conjunction with the units and algorithm steps of the various examples described in the embodiments disclosed herein, the present invention can be implemented in hardware or a combination of hardware and computer software. Whether a function is executed in hardware or by computer software driving hardware depends on the specific application and design constraints of the technical solution. Those skilled in the art may use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of the present invention.

[0072] In an exemplary embodiment, the present invention provides a drone detection device in the form of a virtual device. Figure 5 This is a schematic diagram illustrating the composition of a drone detection device provided in an embodiment of the present invention. Figure 5 As shown, the device includes an acquisition module 501 and a processing module 502.

[0073] The acquisition module 501 is used to acquire the radio frequency signal to be detected; the radio frequency signal is obtained by filtering the acquired original radio frequency signal based on a filter bank; the filter bank includes at least one filter; the at least one filter is used to filter out interference signals from the original radio frequency signal that have different frequency bands and signal bandwidths specified in different frequency bands of the UAV communication protocol; The processing module 502 is used to generate a time-frequency image based on radio frequency signals; to detect the time-frequency image using a preset detection model and generate a detection bounding box of radio frequency signals on the time-frequency image; to correct the time-domain and frequency-domain components of the detection bounding box based on the signal characteristics of the radio frequency signals to obtain a corrected detection bounding box; and to perform UAV detection on an image region selected from the time-frequency image based on the corrected detection bounding box.

[0074] In some possible embodiments, the processing module 502 is specifically used to correct the frequency domain components of the detection bounding box based on the estimated bandwidth of the radio frequency signal; and to correct the time domain components of the detection bounding box based on the Zadov-Zhu ZC sequence of the radio frequency signal.

[0075] In some possible embodiments, the processing module 502 is specifically configured to, if the frequency band Fd of the detected radio frequency signal is not within the specified frequency band of the UAV communication protocol... Within this range, the frequency domain components of the detected bounding box are corrected according to the following formula: ; ; in, Indicates the first A specified frequency band, , express The center frequency; This represents the maximum frequency in the frequency domain component; Represents the minimum frequency in the frequency domain component; This indicates the estimated bandwidth.

[0076] In some possible embodiments, the processing module 502 is further configured to divide the radio frequency signal into segments of length 1. window function Divided into Overlapping sequences: ; in, Indicates the first division Overlapping sequences at position The value at; Indicates radio frequency signal at The value at; Indices representing the overlapping sequences; Indicates the offset. ; based on The power spectrum of each overlapping sequence is calculated using the following formula: ; in, Represents the power spectrum of a radio frequency signal; Indicates angular frequency; The initial estimated bandwidth is determined based on the power spectrum using the following formula: ; in, Indicates the initial estimated bandwidth; express The maximum frequency corresponding to the maximum power; express The minimum frequency corresponding to the maximum power; Based on the initial estimated bandwidth, the bandwidth is estimated using the following formula: ; in, Indicates the total number of subcarriers; This indicates the number of subcarriers used for data transmission.

[0077] In some possible embodiments, the processing module 502 is further configured to perform an autocorrelation calculation on the radio frequency signal according to the following formula:

[0078] in, Indicates the autocorrelation value; Indicates the length of the window; Indicates radio frequency signal; Indicates a time index; Indicates signal The conjugate of complex numbers; Determine the autocorrelation value peak position ; Based on peak position Calculate the total number of subcarriers using the following formula: ; in, Indicates the sampling frequency; This indicates the bandwidth of the radio frequency signal.

[0079] In some possible embodiments, processing module 502 is specifically configured to determine the ZC sequence of the radio frequency signal according to the following formula: ; in, Represents a ZC sequence; Indicates the root index; Indicates the sequence length of the ZC sequence; The ZC sequence and the radio frequency signal are cross-correlated according to the following formula: ; in, This represents the cross-correlation value between the ZC sequence and the radio frequency signal; Indicates the length of the window; The total length of the radio frequency signal; Indicates radio frequency signal at The value; Based on cross-correlation value Peak index The temporal components of the detected bounding box are corrected according to the following formula: ; in, Represents the minimum time in the time-domain component; Indicates the preset coefficient; This indicates the length of the cyclic prefix (CP) under normal circumstances. Indicates the length of the extended loop prefix; This represents the maximum time in the time-domain component; Indicates the specified duration of a broadcast frame. , This indicates the number of symbols in a broadcast frame. Indicates the CP length; Indicates the total number of subcarriers in the radio frequency signal; This indicates the bandwidth of the radio frequency signal.

[0080] In some possible embodiments, processing module 502 is further configured to process cross-correlation values ​​before correcting the temporal components of the detection bounding box. Divide into Q non-overlapping sequences, each of length P, and PQ = LV; Calculate the energy value of each sequence using the following formula: ; in, Indicates energy value; Indicates a sequence index; The interference signal index set is determined according to the following formula. : ; in, This indicates the preset weighting factor. This represents the average energy value of all sequences. right Correct according to the following formula: ; in, This represents the corrected cross-correlation value; This represents the cross-correlation value before correction.

[0081] In some possible embodiments, the processing module 502 is specifically used to obtain a detection radio frequency signal based on the radio frequency signal corresponding to the image region; shift the signal frequency of the detection radio frequency signal from the carrier frequency to near zero frequency to obtain an initial baseband signal; filter the initial baseband signal through a low-pass filter, and then resample the initial baseband signal according to a preset sampling rate to obtain a target baseband signal. ; The autocorrelation value of the target baseband signal over the CP length range is calculated using the following formula: ; in, This represents the autocorrelation value of the target baseband signal within the CP length range; Indicates the CP length; This indicates the length of an OFDM symbol excluding the CP portion; based on The peak index is used to determine the starting index of the OFDM symbol, and the complete OFDM symbol is truncated based on the starting index. The OFDM symbol includes a CP part and a main body part. The time-domain signal of the main body part is extracted, and then transformed to the frequency domain through a Fast Fourier Transform to obtain the frequency domain signal. The phase value of the pilot subcarrier is extracted from the frequency domain signal, and the phase difference between the current symbol and the same pilot subcarrier in the previous symbol is calculated. The frequency offset estimate is calculated using the following formula: ; in, This represents the estimated frequency offset. Indicates the number of points in the Fast Fourier Transform; Indicates the duration of an OFDM symbol; Phase correction is performed on the target baseband signal based on the frequency offset estimate to obtain the corrected target baseband signal; the corrected target baseband signal is used as a bitstream and descrambled based on the Gold sequence to restore the original bitstream; error correction processing is performed on the original bitstream using the Turbo decoder to recover the payload information; UAV detection is performed based on the payload information.

[0082] It should be noted that, Figure 5 The module division shown is illustrative and represents only one logical functional division; in actual implementation, other division methods are possible. For example, two or more functions can be integrated into a single processing module. These integrated modules can be implemented in hardware or as software functional units.

[0083] In an embodiment of the present invention, as described above, the drone detection device may be an electronic device with computing processing capabilities, such as a computer or server. Figure 6 This is a schematic diagram illustrating the composition of an electronic device provided in an embodiment of the present invention. (For example...) Figure 6 As shown, the electronic device includes a processor 602, a communication interface 603, and a bus 604. As an example, the electronic device may also include a memory 601.

[0084] Processor 602 may implement or execute various exemplary logic blocks, modules, and circuits described in connection with the present invention disclosure. Processor 602 may be a central processing unit, a general-purpose processor, a digital signal processor, an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, transistor logic devices, hardware components, or any combination thereof. It may implement or execute various exemplary logic blocks, modules, and circuits described in connection with the present invention disclosure. Processor 602 may also be a combination that implements computational functions, such as including one or more microprocessor combinations, a combination of a DSP and a microprocessor, etc.

[0085] Communication interface 603 is used to connect to other devices via a communication network. This communication network can be Ethernet, wireless access network, wireless local area network (WLAN), etc.

[0086] The memory 601 may be a read-only memory (ROM) or other type of static storage device capable of storing static information and instructions, random access memory (RAM) or other type of dynamic storage device capable of storing information and instructions, or electrically erasable programmable read-only memory (EEPROM), disk storage medium or other magnetic storage device, or any other medium capable of carrying or storing desired program code in the form of instructions or data structures and accessible by a computer, but is not limited thereto.

[0087] In one possible implementation, the memory 601 can exist independently of the processor 602. The memory 601 can be connected to the processor 602 via a bus 604 and is used to store instructions or program code. When the processor 602 calls and executes the instructions or program code stored in the memory 601, it can implement the UAV detection method provided in this embodiment of the invention.

[0088] In another possible implementation, the memory 601 can also be integrated with the processor 602.

[0089] Bus 604 can be an extended industry standard architecture (EISA) bus, etc. Bus 604 can be divided into address bus, data bus, control bus, etc. For ease of representation, Figure 6 The bus is represented by a single thick line, but this does not mean that there is only one bus or one type of bus.

[0090] In exemplary embodiments, this disclosure also provides a readable storage medium including software instructions that, when executed in an electronic device, cause the electronic device to perform the methods described in the above embodiments. The readable storage medium can also be an external storage device of the electronic device, such as a plug-in hard drive, smart media card (SMC), secure digital (SD) card, flash card, etc., equipped on the electronic device. Further, the readable storage medium can include both internal storage units and external storage devices of the electronic device. The readable storage medium is used to store the software instructions and other programs and data required by the electronic device. The readable storage medium can also be used to temporarily store data that has been output or will be output.

[0091] In an exemplary embodiment, this disclosure also provides a computer program product including computer instructions that, when executed on an electronic device, cause the electronic device to perform the methods described in the above method embodiments.

[0092] In the above embodiments, implementation can be achieved, in whole or in part, through software, hardware, firmware, or any combination thereof. When implemented using software programs, implementation can be, in whole or in part, in the form of a computer program product. This computer program product includes one or more computer-executable instructions. When these computer-executable instructions are loaded and executed on a computer, all or part of the flow or function according to the embodiments of the present invention is generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer-executable instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, computer-executable instructions can be transmitted from one website, computer, server, or data center to another via wired (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium accessible to a computer or a data storage device containing one or more servers, data centers, etc., that can be integrated with the medium. The available media can be magnetic media (e.g., floppy disks, hard disks, magnetic tapes), optical media (e.g., DVDs), or semiconductor media (e.g., solid-state disks, SSDs).

[0093] Although the invention has been described herein in conjunction with various embodiments, those skilled in the art will understand and implement other variations of the disclosed embodiments by reviewing the accompanying drawings, the disclosure, and the appended claims in carrying out the claimed invention. In the claims, the word "comprising" does not exclude other components or steps, and "a" or "an" does not exclude a plurality. A single processor or other unit can implement several functions listed in the claims. While different dependent claims may recite certain measures, this does not mean that these measures cannot be combined to produce good results.

[0094] Although the invention has been described in conjunction with specific features and embodiments, it is obvious that various modifications and combinations can be made therein without departing from the spirit and scope of the invention. Accordingly, this specification and drawings are merely exemplary descriptions of the invention as defined by the appended claims, and are considered to cover any and all modifications, variations, combinations, or equivalents within the scope of the invention. Clearly, those skilled in the art can make various alterations and modifications to the invention without departing from its spirit and scope. Thus, if such modifications and modifications of the invention fall within the scope of the claims and their equivalents, the invention is also intended to include such modifications and modifications.

[0095] The above description is merely a preferred embodiment of the present invention and is not intended to limit the technical solution of the present invention in any way. Those skilled in the art should understand that, without departing from the spirit and principles of the present invention, the technical solution can be modified and replaced in several simple ways, and these modifications and replacements are all within the scope of protection covered by the claims.

Claims

1. A method for detecting unmanned aerial vehicles (UAVs), characterized in that, include: Acquire the radio frequency signal to be detected; Based on the radio frequency signal, a time-frequency image is generated; The time-frequency image is detected using a preset detection model, and a detection bounding box for the radio frequency signal is generated on the time-frequency image. Based on the signal characteristics of the radio frequency signal, the time-domain and frequency-domain components of the detection bounding box are corrected to obtain the corrected detection bounding box. UAV detection is performed on the image regions selected from the time-frequency images based on the corrected detection bounding boxes.

2. The UAV detection method according to claim 1, characterized in that, The correction of the time-domain and frequency-domain components of the detection bounding box based on the signal characteristics of the radio frequency signal includes: Based on the estimated bandwidth of the radio frequency signal, the frequency domain components of the detection bounding box are corrected. The temporal components of the detection bounding box are corrected based on the Zadov-Zhu ZC sequence of the radio frequency signal.

3. The UAV detection method according to claim 2, characterized in that, The step of correcting the frequency domain components of the detection bounding box based on the estimated bandwidth of the radio frequency signal includes: To obtain the estimated bandwidth, including: The radio frequency signal is divided into lengths of Window function Divided into Overlapping sequences: ; in, Indicates the first division Overlapping sequences at position The value at; Indicates that the radio frequency signal is in The value at; Indices representing the overlapping sequences; Indicates the offset. ; Based on the above The power spectrum of each overlapping sequence is calculated using the following formula: ; in, This represents the power spectrum of the radio frequency signal; Indicates angular frequency; Based on the power spectrum, the initial estimated bandwidth is determined according to the following formula: ; in, This indicates the initial estimated bandwidth; express The maximum frequency corresponding to the maximum power; express The minimum frequency corresponding to the maximum power; Based on the initial estimated bandwidth, the estimated bandwidth is calculated according to the following formula: ; in, Indicates the total number of subcarriers; Indicates the number of subcarriers used for data transmission; If the frequency band Fd of the detected radio frequency signal is not within the specified frequency band of the UAV communication protocol. Within this range, the frequency domain components of the detected bounding box are corrected according to the following formula: ; ; in, Indicates the first A specified frequency band, , express The center frequency; This represents the maximum frequency in the frequency domain component; This represents the minimum frequency in the frequency domain component; This indicates the estimated bandwidth.

4. The UAV detection method according to claim 2, characterized in that, The Zadov-Zhu ZC sequence based on the radio frequency signal is used to correct the temporal components of the detection bounding box, including: The ZC sequence of the radio frequency signal is determined according to the following formula: ; in, This represents the ZC sequence; Indicates the root index; Indicates the sequence length of the ZC sequence; The ZC sequence and the radio frequency signal are cross-correlated according to the following formula: ； in, This represents the cross-correlation value between the ZC sequence and the radio frequency signal; Indicates the length of the window; The total length of the radio frequency signal; Indicates that the radio frequency signal is in The value; Based on the cross-correlation value Peak index The temporal components of the detected bounding box are corrected according to the following formula: ; in, This represents the minimum time in the time-domain component; Indicates the preset coefficient; This indicates the length of the cyclic prefix (CP) under normal circumstances. Indicates the length of the extended loop prefix; This represents the maximum time in the time-domain component; Indicates the specified duration of a broadcast frame. , This indicates the number of symbols in a broadcast frame. Indicates the CP length; This represents the total number of subcarriers of the radio frequency signal; This indicates the bandwidth of the radio frequency signal.

5. The UAV detection method according to claim 2, characterized in that, The Zadov-Zhu ZC sequence based on the radio frequency signal is used to correct the temporal components of the detection bounding box, including: The ZC sequence of the radio frequency signal is determined according to the following formula: ; in, This represents the ZC sequence; Indicates the root index; Indicates the sequence length of the ZC sequence; The ZC sequence and the radio frequency signal are cross-correlated according to the following formula: ； in, This represents the cross-correlation value between the ZC sequence and the radio frequency signal; Indicates the length of the window; The total length of the radio frequency signal; Indicates that the radio frequency signal is in The value; Based on the cross-correlation value Peak index The temporal components of the detected bounding box are corrected according to the following formula: ; in, This represents the minimum time in the time-domain component; Indicates the preset coefficient; This indicates the length of the cyclic prefix (CP) under normal circumstances. Indicates the length of the extended loop prefix; This represents the maximum time in the time-domain component; Indicates the specified duration of a broadcast frame. , This indicates the number of symbols in a broadcast frame. Indicates the CP length; This represents the total number of subcarriers of the radio frequency signal; This indicates the bandwidth of the radio frequency signal.

6. The UAV detection method according to claim 5, characterized in that, cross-correlation value Make corrections, including: cross-correlation value Divide into Q non-overlapping sequences, each of length P, and PQ = LV; Calculate the energy value of each sequence using the following formula: ; in, Indicates energy value; Indicates a sequence index; The interference signal index set is determined according to the following formula. : ； in, This indicates the preset weighting factor. This represents the average energy value of all sequences. right Correct according to the following formula: ; in, This represents the corrected cross-correlation value; This represents the cross-correlation value before correction.

7. The UAV detection method according to claim 1, characterized in that, The process of detecting unmanned aerial vehicles (UAVs) by selecting image regions from the time-frequency image based on the corrected detection bounding box includes: Based on the radio frequency signal corresponding to the image region, a detection radio frequency signal is obtained; The frequency of the detected radio frequency signal is shifted from the carrier frequency to near zero frequency to obtain the initial baseband signal; After filtering the initial baseband signal using a low-pass filter, the initial baseband signal is resampled according to a preset sampling rate to obtain the target baseband signal. ; The autocorrelation value of the target baseband signal over the CP length range is calculated using the following formula: ; in, This represents the autocorrelation value of the target baseband signal within the CP length range; Indicates the CP length; This indicates the length of an OFDM symbol excluding the CP portion; based on The peak index is used to determine the starting index of the OFDM symbol, and the complete OFDM symbol is truncated based on the starting index; the OFDM symbol includes a CP part and a main body part; The time-domain signal of the main body is extracted and transformed into the frequency domain through a fast Fourier transform to obtain the frequency domain signal; The phase value of the pilot subcarrier is extracted from the frequency domain signal, and the phase difference between the current symbol and the same pilot subcarrier in the previous symbol is calculated. ; The frequency shift estimate is calculated using the following formula: ； in, This represents the estimated frequency offset. Indicates the number of points in the Fast Fourier Transform; Indicates the duration of an OFDM symbol; Based on the frequency offset estimate, the target baseband signal is phase-corrected to obtain the corrected target baseband signal; The corrected target baseband signal is used as a bitstream and descrambled based on the Gold sequence to restore the original bitstream. The original bitstream is corrected using a Turbo decoder to recover the payload information; UAV detection is performed based on the payload information.

8. An electronic device, characterized in that, include: Processor and memory; The memory stores instructions that the processor can execute; When the processor is configured to execute the instructions, the electronic device performs the method as described in any one of claims 1-7.

9. A readable storage medium, characterized in that, include: Software instructions; When the software instructions are executed in an electronic device, the electronic device causes the electronic device to perform the method as described in any one of claims 1-7.

10. A computer program product, characterized in that, include: Computer instructions; When the computer instructions are executed in an electronic device, the electronic device causes the electronic device to perform the method as described in any one of claims 1-7.

Images