UAV accurate identification methods, equipment, storage media and products

Abstract

The present invention discloses a method, device, storage medium, and product for accurately identifying drones. The method includes performing Fourier transform on the drone's image transmission signal to obtain a time-frequency diagram; preprocessing and segmenting the image transmission signal according to the time-frequency diagram to obtain multiple baseband signal segments containing preambles or pilot sequences; determining the subcarriers of the OFDM symbols of the preamble or pilot sequence and the subcarriers carrying data in each baseband signal segment, and generating a local sequence for each baseband signal segment; calculating the normalized correlation sequence between each baseband signal segment and each of its local sequences, and then determining the generation method of the preamble or pilot sequence in each baseband signal segment; and accurately identifying the drone based on the generation method of the preamble or pilot sequence in all baseband signal segments. The present invention achieves accurate identification of drones without requiring knowledge of the drone's unique serial number.

Description

Technical Field

[0001] The present invention belongs to the field of drone detection technology, and in particular relates to a method, device, storage medium and product for accurately identifying drones. Background Art

[0002] Currently, there are many spectrum detection devices on the market. These devices are generally passive and use algorithms to detect and identify drones. However, when a spectrum detection device detects a drone, it is usually impossible to directly determine whether it has been detected before. Because a drone may appear within the spectrum multiple times, and each drone signal may be affected by environmental factors, flight conditions, interference, and other factors, accurate and consistent identification cannot be guaranteed.

[0003] With the increasing number of drones in low-altitude airspace, the Civil Aviation Administration of China has introduced a management system to prevent conflicts between drones and other aircraft and ensure safety in low-altitude airspace. These regulations require drones to be registered with a unique serial number upon production, report the serial number and flight data, and restrict no-fly zones. Although drones have fixed identity information, not all organizations have access to it. Therefore, the indiscriminate use of drones can still lead to safety incidents and privacy breaches.

[0004] To accurately identify drones, some researchers are attempting to decode the WiFi or broadcast signals by cracking their protocols in an attempt to extract their original serial numbers. However, with so many drone manufacturers and types, not all drones contain broadcast signals, WiFi signals, or unique serial numbers, especially lightweight ones. This makes it difficult to accurately identify drones without their unique serial numbers. Summary of the Invention

[0005] The purpose of the present invention is to provide a method, device, storage medium and product for accurately identifying drones, so as to solve the problem of being unable to accurately identify whether a drone has been detected by a detection device before when the drone's unique serial number is unknown.

[0006] The present invention solves the above technical problems through the following technical solutions: a method for accurately identifying drones, comprising:

[0007] Get the image transmission signal of the drone;

[0008] Performing Fourier transform on the image transmission signal to obtain a time-frequency graph;

[0009] Preprocessing and segmenting the image transmission signal according to the time-frequency diagram to obtain multiple baseband signal segments containing preamble codes or pilot sequences;

[0010] Determining, based on the time-frequency diagram, the subcarriers of the OFDM symbols of the preamble or pilot sequence and the subcarriers carrying data in each baseband signal segment;

[0011] Generate a local sequence for each baseband signal segment based on the subcarriers of the OFDM symbols of the preamble or pilot sequence, the subcarriers carrying data, and the carried data in each baseband signal segment;

[0012] Calculating a normalized correlation sequence between each baseband signal segment and each of its local sequences;

[0013] determining a method for generating a preamble or pilot sequence in each baseband signal segment according to a normalized correlation sequence between each baseband signal segment and each local sequence thereof;

[0014] Accurate identification of drones is achieved based on the generation method of preamble codes or pilot sequences in all baseband signal segments.

[0015] Furthermore, the image transmission signal is preprocessed and segmented according to the time-frequency graph, specifically including:

[0016] Extracting key information of the image transmission signal from the time-frequency diagram; wherein the key information includes bandwidth, frequency band, center frequency, data starting point and data end point;

[0017] Down-converting the image transmission signal according to the frequency band and center frequency;

[0018] Designing a filter according to the bandwidth, and using the filter to filter the down-converted image transmission signal;

[0019] Perform time domain decimation on the filtered image transmission signal to obtain a baseband signal; the decimation factor is determined by the sampling rate and bandwidth;

[0020] According to the data starting point and the extraction multiple, the baseband signal is evenly divided according to the time window to obtain multiple baseband signal segments containing preamble codes or pilot sequences; wherein the length of the time window is the interval between two pilot sequences in the baseband signal.

[0021] Furthermore, a pre-trained deep learning model is used to extract key information of the image transmission signal from the time-frequency graph. The training process of the deep learning model includes:

[0022] Constructing a sample data set; wherein the samples in the sample data set include a time-frequency graph and key information thereof;

[0023] Deploy a deep learning model and use the sample dataset to train, verify, and evaluate the deep learning model.

[0024] Furthermore, determining the subcarriers of the OFDM symbols of the preamble or pilot sequence and the subcarriers carrying data in each baseband signal segment according to the time-frequency diagram specifically includes:

[0025] determining a position of a preamble or a pilot sequence in each baseband signal segment according to the time-frequency diagram;

[0026] extracting the preamble or pilot sequence in each baseband signal segment based on a position of the preamble or pilot sequence in each baseband signal segment;

[0027] Performing Fourier transform on the preamble or pilot sequence in each baseband signal segment to obtain subcarriers of the OFDM symbol of the preamble or pilot sequence in each baseband signal segment;

[0028] The subcarriers carrying data are determined according to the subcarriers of the OFDM symbols of the preamble or pilot sequence in each baseband signal segment.

[0029] Furthermore, the specific calculation formula for the normalized correlation sequence between each baseband signal segment and each of its local sequences is:

[0030] ;

[0031] ;

[0032] in, Indicates the baseband signal segment With its local sequence The normalized correlation sequence of Indicates the baseband signal segment With its local sequence The related sequence of Represents the baseband signal segment, represents the local sequence of baseband signal segments, Indicates delay, represents the index of discrete time, Indicates the baseband signal segment length, Represents a local sequence length.

[0033] Furthermore, determining a method for generating a preamble or pilot sequence in each baseband signal segment based on a normalized correlation sequence between each baseband signal segment and each local sequence thereof specifically includes:

[0034] For each baseband signal segment, taking the maximum normalized correlation value in each normalized correlation sequence, and comparing the maximum normalized correlation value with a set threshold; if the maximum normalized correlation value exceeds the set threshold, recording the maximum normalized correlation value and the local sequence corresponding to the maximum normalized correlation value; otherwise, discarding the normalized correlation sequence;

[0035] If multiple maximum normalized correlation values and their corresponding local sequences are recorded, the largest one is selected from the multiple maximum normalized correlation values;

[0036] Determine, according to the local sequence corresponding to the largest one, the subcarriers of the OFDM symbol of the preamble or pilot sequence in the baseband signal segment, the subcarriers carrying data, and the data carried by the subcarriers;

[0037] The generation method of the preamble code or pilot sequence in the baseband signal segment is determined according to the subcarriers of the OFDM symbols of the preamble code or pilot sequence in the baseband signal segment, the subcarriers carrying data, and the data carried by the subcarriers.

[0038] Based on the same concept, the present invention also provides an electronic device, including a memory, a processor, and a computer program / instruction stored in the memory, wherein the processor executes the computer program / instruction to implement the above-mentioned method for accurately identifying drones.

[0039] Based on the same concept, the present invention also provides a computer-readable storage medium having a computer program / instruction stored thereon, which, when executed by a processor, implements the above-mentioned method for accurately identifying drones.

[0040] Based on the same concept, the present invention also provides a computer program product, including a computer program / instruction, which implements the above-mentioned drone accurate identification method when executed by a processor.

[0041] Compared with the prior art, the advantages of the present invention are:

[0042] The present invention analyzes the image transmission signal to determine the generation method of the preamble code or pilot sequence in each baseband signal segment, and then determines the type of generation method of the preamble code or pilot sequence in the entire baseband signal. Combined with the characteristic that the types of generation methods of the preamble code or pilot sequence in the baseband signals of different drones are not exactly the same, it can accurately identify whether the drone has been detected by the detection equipment before. The method of the present invention, in conjunction with the spectrum detection equipment, can accurately identify the drone and whether the drone has been detected before, ensuring the accuracy and uniqueness of the identification result, and achieving accurate identification of the drone without knowing the drone's unique serial number.

[0043] The present invention achieves accurate drone identification by analyzing the image transmission signal to obtain the type of generation method of the preamble code or pilot sequence in the baseband signal, thereby improving the distance of drone identification and realizing accurate long-distance drone identification; the identification process of the present invention is relatively simple, with a small amount of calculation, and can be processed in parallel with the spectrum detection equipment to identify drones, thereby shortening the identification time. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] In order to more clearly illustrate the technical solution of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only one embodiment of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0045] Figure 1 This is a flow chart of the method for accurately identifying drones in an embodiment of the present invention. DETAILED DESCRIPTION

[0046] The following is a clear and complete description of the technical solutions of the present invention in conjunction with the accompanying drawings of the embodiments of the present invention. Obviously, the embodiments described are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts are within the scope of protection of the present invention.

[0047] The following specific embodiments are used to describe the technical solution of the present application in detail. The following specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described in detail in some embodiments.

[0048] Example 1

[0049] In order to solve the problem of being unable to accurately identify whether a drone has been detected by a detection device before when the unique serial number of the drone cannot be known, the present invention provides a method for accurately identifying drones. The method determines the generation method of the preamble code or pilot sequence in each baseband signal segment by analyzing the image transmission signal. Based on the characteristic that the types of generation methods of the preamble codes or pilot sequences in the baseband signals of different drones are not exactly the same, the method accurately identifies whether a drone has been detected by a detection device before.

[0050] Figure 1 The flowchart of the method for accurately identifying drones in an embodiment of the present invention is shown in FIG. Figure 1 As shown, the method for accurately identifying drones includes the following steps:

[0051] Step S1: Obtain the image transmission signal of the drone.

[0052] Drone communication links are divided into uplinks and downlinks. The uplink is used for ground-to-drone communication, primarily transmitting drone flight control data and sensor control data. The downlink is used for drone-to-ground communication, primarily transmitting navigation information such as the drone's path, speed, and distance. In this embodiment, the drone's image transmission signal is a downlink signal.

[0053] When a drone is operating, the drone detection device performs passive detection. Upon detecting a drone, the device saves the image transmission signal. The real and imaginary components of the image transmission signal are stored separately. When used, the real and imaginary components are combined into a complex signal. The image transmission signal in this embodiment is a complex signal.

[0054] Step S2: Perform Fourier transform on the image transmission signal to obtain a time-frequency diagram.

[0055] When performing a Fourier transform, the number of Fourier transform points (i.e., the number of FFT points) can be selected. Typically, to improve computational efficiency, the number of FFT points can be selected as a power of 2. In this embodiment, 1024 data points are selected as the number of points for a single Fourier transform, i.e., the number of FFT points is 1024.

[0056] Step S3: Preprocess and segment the image transmission signal according to the time-frequency diagram to obtain multiple baseband signal segments containing preamble codes or pilot sequences.

[0057] In a specific embodiment of the present invention, the image transmission signal is preprocessed and segmented according to the time-frequency diagram, specifically including:

[0058] Step S3.1: Extract key information of the image transmission signal from the time-frequency diagram; wherein the key information includes the bandwidth, frequency band, center frequency, data starting point, and data end point of the image transmission signal.

[0059] In this embodiment, a pre-trained deep learning model is used to extract key information of the image transmission signal from the time-frequency graph. The training process of the deep learning model includes:

[0060] Step S3.11: Construct a sample dataset.

[0061] Collect time-frequency plots of image transmission signals from different types of drones, including but not limited to those from different manufacturers, models, frequency bands, modulation methods, and power levels. Manually annotate the collected time-frequency plots with information such as bandwidth, frequency band, center frequency, data start and end points. Construct a sample dataset based on the time-frequency plots and their annotations. Each sample in the sample dataset includes the time-frequency plot, its bandwidth, frequency band, center frequency, data start and end points, and more.

[0062] The sample data set is divided into a training set and a validation set. The training set is used to train the deep learning model, and the validation set is used to validate the deep learning model.

[0063] Step S3.12: Deploy the deep learning model and use the sample dataset to train, verify, and evaluate the deep learning model.

[0064] Based on the characteristics of the sample data and the effectiveness of extracting key information, select an appropriate deep learning model, such as a hybrid network of convolutional neural networks and long-short-term memory, or a hybrid network of convolutional neural networks and recurrent neural networks. The deep learning model is trained using the time-frequency graph as input and the corresponding annotations as the expected output. During training, the mean square error (MSE) loss function is used to calculate the loss error, and the parameters of the deep learning model are adjusted based on the loss error to achieve training. If the loss error does not decrease significantly over multiple consecutive training cycles, or if the set number of training cycles is reached, training is terminated to avoid overfitting.

[0065] Evaluate the trained and verified deep learning model, and based on the evaluation results, determine whether it is necessary to increase the sample data set and continue training until the performance evaluation requirements are met.

[0066] Step S3.2: Down-convert the image transmission signal according to the frequency band and center frequency.

[0067] In order to observe the down-conversion process more clearly in the frequency domain, the down-converted image transmission signal is subjected to Fourier transform to obtain the corresponding time-frequency diagram.

[0068] Step S3.3: Design a filter according to the bandwidth, and use the filter to filter the down-converted image transmission signal.

[0069] Different filter coefficients are set based on the bandwidth to design the corresponding filter. The filter is then used to low-pass filter the down-converted image transmission signal to remove out-of-band interference. To better observe the filtering effect, a Fourier transform is performed on the filtered image transmission signal to generate the corresponding time-frequency plot.

[0070] Step S3.4: Perform time domain extraction on the filtered image transmission signal to obtain a baseband signal.

[0071] The decimation factor is determined based on the sampling rate and bandwidth, and the filtered image transmission signal is subjected to time-domain decimation to reduce the computational complexity of the correlation sequence. To better observe the time-frequency decimation effect, the baseband signal is Fourier transformed to obtain the corresponding time-frequency plot.

[0072] Step S3.5: According to the data starting point and the decimation multiple, the baseband signal is evenly divided according to the time window to obtain multiple baseband signal segments containing preamble codes or pilot sequences.

[0073] The length of the time window is the interval between two pilot sequences in the baseband signal, and the length of the baseband signal segment is equal to the length of the time window. For example, assuming the data starting point of the deep learning model output image transmission signal is 1000, the interval between pilot sequences in the baseband signal is 2000, and the decimation factor is 2, then the data starting point of the baseband signal is 500. Starting from the 500th data point, the baseband signal is evenly divided into segments of 2000 data points each to obtain baseband signal segments.

[0074] Step S4: Determine the subcarriers of the OFDM (Orthogonal Frequency Division Multiplexing) symbols of the preamble or pilot sequence and the subcarriers carrying data in each baseband signal segment according to the time-frequency diagram.

[0075] In a specific embodiment of the present invention, determining the subcarriers of the OFDM symbols of the preamble or pilot sequence and the subcarriers carrying data in each baseband signal segment according to the time-frequency diagram specifically includes:

[0076] Step S4.1: Determine the position of the preamble or pilot sequence in each baseband signal segment according to the time-frequency diagram;

[0077] Step S4.2: extracting the preamble or pilot sequence in each baseband signal segment based on the position of the preamble or pilot sequence in each baseband signal segment;

[0078] Step S4.3: Performing Fourier transform on the preamble or pilot sequence in each baseband signal segment to obtain the subcarriers of the OFDM symbol of the preamble or pilot sequence in each baseband signal segment;

[0079] Step S4.4: Determine the subcarriers carrying data according to the subcarriers of the OFDM symbols of the preamble or pilot sequence in each baseband signal segment.

[0080] The subcarriers carrying data refer to which subcarriers can carry data. For example, the number of subcarriers in an OFDM symbol is 1024, of which 150 subcarriers carry data.

[0081] Step S5: Generate a local sequence for each baseband signal segment according to the subcarriers of the OFDM symbols of the preamble or pilot sequence in each baseband signal segment, the subcarriers carrying data, and the carried data.

[0082] The data carried is a type of sequence whose autocorrelation characteristics are optimized through mathematical construction. The specific data is multiple and definite, and the length and type of the data are the same as the number of subcarriers carrying the data.

[0083] Define all possible data carrying methods, each corresponding to a set of subcarriers carrying data. Combine the subcarriers of the OFDM symbol with different data carrying methods and the carried data to generate different local sequences.

[0084] For example, if the number of subcarriers in the OFDM symbol of the pilot sequence is 1024 and the number of subcarriers carrying data is 150, then 150 possible data-carrying subcarriers are selected from the 1024 subcarriers. Each selected group (i.e., 150 data-carrying subcarriers) generates a data-carrying pattern. Combining the 1024 subcarriers with a group of subcarriers and a type of carried data generates a local sequence. Therefore, the number of local sequences equals the product of the number of data-carrying patterns and the type of carried data.

[0085] Step S6: Calculate the normalized correlation sequence between each baseband signal segment and each of its local sequences.

[0086] In this embodiment, the calculation formula for the correlation sequence between each baseband signal segment and each of its local sequences is:

[0087] (1)

[0088] in, Indicates the baseband signal segment With its local sequence The related sequence of Represents the baseband signal segment, represents the local sequence of baseband signal segments, Indicates delay, represents the index of discrete time, Indicates the baseband signal segment length, Represents a local sequence length.

[0089] In order to eliminate the influence of signal length and signal amplitude on the correlation value, the correlation sequence is normalized to obtain a normalized correlation sequence. The normalization formula is:

[0090] (2)

[0091] in, Indicates the baseband signal segment With its local sequence The normalized correlation sequence of .

[0092] Step S7: determining a generation method of a preamble code or pilot sequence in each baseband signal segment according to a normalized correlation sequence between each baseband signal segment and each local sequence thereof.

[0093] In a specific embodiment of the present invention, a method for generating a preamble or pilot sequence in each baseband signal segment is determined based on a normalized correlation sequence between each baseband signal segment and each of its local sequences, specifically including:

[0094] Step S7.1: For each baseband signal segment, take each normalized correlation sequence The maximum normalized correlation value in ,in, represents the normalized correlation sequence between the baseband signal segment and its i-th local sequence The maximum normalized correlation value in .

[0095] Step S7.2: The maximum normalized correlation value Compare with the set threshold;

[0096] If the maximum normalized correlation value If the set threshold is exceeded, the maximum normalized correlation value is recorded. and the maximum normalized correlation value Corresponding local sequence ; Otherwise, discard the normalized correlation sequence .

[0097] For the baseband signal segment Repeat steps S7.1 and S7.2 for each normalized correlation sequence to obtain all the maximum normalized correlation values recorded. and the local sequence corresponding to the maximum normalized correlation value The threshold value is set based on experience, and in this embodiment, the threshold value is set to 0.2.

[0098] Step S7.3: If there are multiple maximum normalized correlation values and its corresponding local sequence are recorded, then the largest one is selected from multiple maximum normalized correlation values, that is, the maximum value is max (all the maximum normalized correlation values recorded ).

[0099] Step S7.4: Determine the subcarriers of the OFDM symbols of the preamble or pilot sequence in the baseband signal segment, the subcarriers carrying data, and the data carried by the subcarriers according to the local sequence corresponding to the largest one.

[0100] The local sequence corresponding to the largest value (i.e., the determined local sequence) has the highest similarity to the preamble or pilot sequence in the baseband signal segment. Therefore, the subcarriers of the OFDM symbols of the preamble or pilot sequence in the baseband signal segment, the subcarriers carrying data, and the data carried by the subcarriers can be determined based on the local sequence corresponding to the largest value.

[0101] Step S7.5: Determine the generation method of the preamble code or pilot sequence in the baseband signal segment according to the subcarriers of the OFDM symbols of the preamble code or pilot sequence in the baseband signal segment, the subcarriers carrying data, and the data carried by the subcarriers.

[0102] The preamble code or pilot sequence in the baseband signal segment is generated according to the subcarriers of the OFDM symbol of the preamble code or pilot sequence, the subcarriers carrying data, and the data carried by the subcarriers.

[0103] Step S8: Accurately identify the drone based on the generation method of the preamble code or pilot sequence in all baseband signal segments.

[0104] For each baseband signal segment, the method for generating its preamble or pilot sequence can be determined, and thus the type of preamble or pilot sequence generation method in the entire baseband signal can be determined. The types of preamble or pilot sequence generation methods for baseband signals of different drones may vary. That is, the types of preamble or pilot sequence generation methods for baseband signals of different drones may be partially the same, partially different, or completely different. However, the types of preamble or pilot sequence generation methods for baseband signals of different drones cannot be completely the same.

[0105] Based on the fact that the types of generation methods of the preamble codes or pilot sequences in the baseband signals of different drones are not exactly the same, it is possible to determine whether the drone has been detected by the detection equipment before based on the type of generation method of the preamble code or pilot sequence in the entire baseband signal. This achieves accurate identification of drones when the unique serial number of the drone cannot be known, and can assist detection equipment in drone identification, avoiding the problem that traditional detection equipment cannot directly determine whether the drone has been detected before when detecting the drone.

[0106] Example 2

[0107] An embodiment of the present invention further provides an electronic device, comprising: a memory, a processor, and a computer program / instructions stored on the memory, wherein the processor executes the computer program / instructions to implement the drone accurate identification method in the embodiment of the present application.

[0108] Although not shown, the electronic device includes a processor that can perform various appropriate operations and processes based on programs and / or data stored in a read-only memory (ROM) or programs and / or data loaded from a storage portion into a random access memory (RAM). The processor can be a multi-core processor or can include multiple processors. In some embodiments, the processor can include a general-purpose main processor and one or more special coprocessors, such as a central processing unit, a graphics processing unit (GPU), a neural network processor (NPU), a digital signal processor (DSP), etc. Various programs and data required for device operation are also stored in the RAM. The processor, ROM, and RAM are connected to each other via a bus. An input / output (I / O) interface is also connected to the bus.

[0109] The processor and memory are used together to execute the program / instructions stored in the memory. When the program / instructions are executed by the computer, the methods, steps or functions described in the above embodiments can be implemented.

[0110] Although not shown, an embodiment of the present invention further provides a computer-readable storage medium having a computer program / instruction stored thereon, which, when executed by a processor, implements the method for accurately identifying a drone in an embodiment of the present application.

[0111] Computer-readable storage media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. The information can be computer-readable instructions, data structures, program modules, or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disc read-only memory (CD-ROM), digital versatile disc (DVD) or other optical storage, magnetic cassettes, magnetic disk storage or other magnetic storage devices, or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer-readable media does not include transitory media such as modulated data signals and carrier waves.

[0112] Although not shown, an embodiment of the present invention further provides a computer program product, including: a computer program / instruction, which, when executed by a processor, implements the drone accurate identification method in the embodiment of the present application.

[0113] The above disclosure is only a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any technician familiar with this technical field can easily think of changes or modifications within the technical scope disclosed in the present invention, and they should all be covered by the scope of protection of the present invention.

Claims

1. A method for accurately identifying drones, characterized in that: The identification method comprises: Get the image transmission signal of the drone; Performing Fourier transform on the image transmission signal to obtain a time-frequency graph; Preprocessing and segmenting the image transmission signal according to the time-frequency diagram to obtain multiple baseband signal segments containing preamble codes or pilot sequences; Determining, based on the time-frequency diagram, the subcarriers of the OFDM symbols of the preamble or pilot sequence and the subcarriers carrying data in each baseband signal segment; Generating a local sequence for each baseband signal segment based on the subcarriers of the OFDM symbols of the preamble or pilot sequence, the subcarriers carrying data, and the carried data in each baseband signal segment; wherein different local sequences are generated by combining the subcarriers of the OFDM symbols of the preamble or pilot sequence in each baseband signal segment with different data carrying modes and carried data; Calculating a normalized correlation sequence between each baseband signal segment and each of its local sequences; determining a method for generating a preamble or pilot sequence in each baseband signal segment according to a normalized correlation sequence between each baseband signal segment and each local sequence thereof; Accurately identify drones based on how the preamble or pilot sequence is generated in all baseband signal segments; The method of generating a preamble or pilot sequence in each baseband signal segment is determined according to a normalized correlation sequence between each baseband signal segment and each local sequence thereof, specifically including: For each baseband signal segment, taking the maximum normalized correlation value in each normalized correlation sequence, and comparing the maximum normalized correlation value with a set threshold; if the maximum normalized correlation value exceeds the set threshold, recording the maximum normalized correlation value and the local sequence corresponding to the maximum normalized correlation value; otherwise, discarding the normalized correlation sequence; If multiple maximum normalized correlation values and their corresponding local sequences are recorded, the largest one is selected from the multiple maximum normalized correlation values; Determine, according to the local sequence corresponding to the largest one, the subcarriers of the OFDM symbol of the preamble or pilot sequence in the baseband signal segment, the subcarriers carrying data, and the data carried by the subcarriers; The generation method of the preamble code or pilot sequence in the baseband signal segment is determined according to the subcarriers of the OFDM symbols of the preamble code or pilot sequence in the baseband signal segment, the subcarriers carrying data, and the data carried by the subcarriers.

2. The method for accurately identifying drones according to claim 1, characterized in that: Preprocessing and segmenting the image transmission signal according to the time-frequency graph specifically includes: Extracting key information of the image transmission signal from the time-frequency diagram; wherein the key information includes bandwidth, frequency band, center frequency, data starting point and data end point; Down-converting the image transmission signal according to the frequency band and center frequency; Designing a filter according to the bandwidth, and using the filter to filter the down-converted image transmission signal; Perform time domain decimation on the filtered image transmission signal to obtain a baseband signal; the decimation factor is determined by the sampling rate and bandwidth; According to the data starting point and the extraction multiple, the baseband signal is evenly divided according to the time window to obtain multiple baseband signal segments containing preamble codes or pilot sequences; wherein the length of the time window is the interval between two pilot sequences in the baseband signal.

3. The method for accurately identifying drones according to claim 2, characterized in that: A pre-trained deep learning model is used to extract key information of the image transmission signal from the time-frequency graph. The training process of the deep learning model includes: Constructing a sample data set; wherein the samples in the sample data set include a time-frequency graph and key information thereof; Deploy a deep learning model and use the sample dataset to train, verify, and evaluate the deep learning model.

4. The method for accurately identifying drones according to claim 1, wherein: Determining, according to the time-frequency diagram, the subcarriers of the OFDM symbols of the preamble or pilot sequence and the subcarriers carrying data in each baseband signal segment, specifically comprising: determining a position of a preamble or a pilot sequence in each baseband signal segment according to the time-frequency diagram; extracting the preamble or pilot sequence in each baseband signal segment based on a position of the preamble or pilot sequence in each baseband signal segment; Performing Fourier transform on the preamble or pilot sequence in each baseband signal segment to obtain subcarriers of the OFDM symbol of the preamble or pilot sequence in each baseband signal segment; The subcarriers carrying data are determined according to the subcarriers of the OFDM symbols of the preamble or pilot sequence in each baseband signal segment.

5. The method for accurately identifying drones according to claim 1, wherein: The specific calculation formula for the normalized correlation sequence between each baseband signal segment and each of its local sequences is: ； ； in, Indicates the baseband signal segment With its local sequence The normalized correlation sequence of Indicates the baseband signal segment With its local sequence The related sequence of Represents the baseband signal segment, represents the local sequence of baseband signal segments, Indicates delay, represents the index of discrete time, Indicates the baseband signal segment length, Represents a local sequence length.

6. An electronic device comprising a memory, a processor, and a computer program / instruction stored in the memory, characterized in that: The processor executes the computer program / instructions to implement the method for accurately identifying a drone according to any one of claims 1 to 5.

7. A computer-readable storage medium having a computer program / instruction stored thereon, characterized in that: When the computer program / instruction is executed by a processor, the method for accurately identifying a drone according to any one of claims 1 to 5 is implemented.

8. A computer program product comprising a computer program / instructions, characterized in that When the computer program / instruction is executed by a processor, the method for accurately identifying a drone according to any one of claims 1 to 5 is implemented.

Images