A 915MHz frequency band drone detection method, system and terminal

Abstract

The present application relates to a method, system, and terminal for detecting drones in the 915MHz frequency band. The method comprises: dividing a received signal to be detected into multiple frames of time-domain data according to a first division duration; converting the time-domain data into frequency-domain data; dividing each frame of frequency-domain data into multiple frequency-domain channels; calculating the channel power of a single frequency-domain channel; calculating the average channel power of each frequency-domain channel in a set range of #imgabs0# frames, thereby obtaining a denoised signal power spectrum corresponding to each frame, thereby obtaining a detection threshold for each frame of frequency-domain data, and thereby extracting the corresponding candidate channels from each frame of frequency-domain data; sequentially determining whether the candidate channels in each frame are adjacent; if so, calculating the number of adjacent candidate channels in each frame #imgabs1#; after #imgabs2# meets the number requirement, determining whether the number of occurrences of the corresponding frames meeting the number requirement meets the occurrence requirement; if so, determining that the signal to be detected is a drone signal. The present application has the beneficial effects of shortening the detection time interval and reducing implementation costs.

Description

Technical Field

[0001] The present application relates to the technical field of drones, and in particular to a method, system, and terminal for detecting drones in the 915MHz frequency band. Background Art

[0002] With the widespread adoption of drone technology, drone signals are increasing across various frequency bands. The 915MHz band, in particular, is the preferred choice for many high-performance drones and FPV (first-person view) systems due to its long transmission range and strong penetration, such as those from brands like TBS (Black Sheep). This phenomenon poses serious security risks, particularly in critical areas such as airports, energy facilities, and large-scale event venues, where it can pose a threat to safety and security, while also posing a risk of infringing personal privacy. Therefore, there is an urgent need to develop efficient and accurate drone detection technology to address these increasingly severe security challenges.

[0003] Existing technologies for detecting drones in the 915MHz band are limited. Due to the wide bandwidth and long duration of drone image transmission signals, most solutions focus on detecting drones across the entire frequency band (e.g., 300MHz-6GHz), as outlined in patents CN112147704, CN114095932B, and CN213661639U. Some solutions employ active detection, such as radar-based detection (CN218099589U) and detection based on 4G or 5G communication signals (CN115134037B). These methods require simultaneous detection across multiple frequency bands, resulting in high computational complexity, complex processes, and reliance on specialized hardware. For drones operating in the 915MHz band, the detection intervals are excessively long, leading to high implementation costs. Summary of the Invention

[0004] In order to shorten the detection time interval and reduce the implementation cost, the present application provides a 915MHz frequency band UAV detection method, system and terminal.

[0005] In the first aspect, the present application provides a method for detecting drones in the 915MHz frequency band, which adopts the following technical solutions:

[0006] A method for detecting drones in the 915 MHz frequency band, comprising:

[0007] receiving a signal to be detected;

[0008] Dividing the signal to be detected into multiple frames of time domain data according to a preset first division time length;

[0009] Perform FFT (Fast Fourier Transform) processing on each frame of time domain data to convert it into frequency domain data;

[0010] Divide each frame of frequency domain data into multiple frequency domain channels;

[0011] Calculating the channel power of a single frequency domain channel;

[0012] Calculate the set range an average channel power of each of the frequency domain channels in the frequency domain data of the frame;

[0013] Obtaining a denoised signal power spectrum value corresponding to each frame of the frequency domain data according to each of the average channel powers;

[0014] Obtaining a detection threshold for each frame of the frequency domain data according to the power spectrum value of the denoised signal;

[0015] extracting corresponding candidate channels from each frame of the frequency domain data according to the detection threshold;

[0016] determining in sequence whether there is an adjacent candidate channel in each frame of the frequency domain data;

[0017] If so, calculate the number of adjacent candidate channels in the frequency domain data of the corresponding frame of the adjacent candidate channels. ;

[0018] Judge the Whether the quantity requirements are met;

[0019] If so, determining whether the number of occurrences of the frequency domain data in the corresponding frame where the number of candidate channels meets the number requirement meets the occurrence requirement;

[0020] If so, it is determined that the signal to be detected is a drone signal of a drone in the 915MHz frequency band.

[0021] By adopting the above technical solution, the received signal of a certain time length is regarded as a frame of data, and FFT (Fast Fourier Transform) is performed to transform the time domain of each frame signal into the frequency domain; then, each frame of frequency domain data is divided into several frequency domain channels, and the average signal power of all frequency domain channels is calculated; then, according to each average channel power, the corresponding denoised signal power spectrum is obtained to further calculate the detection threshold of each frame of frequency domain data; then, the corresponding candidate channels in each frame of frequency domain data are extracted, and whether the candidate channels in each frame are adjacent is determined in turn. If adjacent, the number of adjacent candidate channels in each frame of frequency domain data is calculated. ,like If the number requirements are met, it means that the signal to be detected meets the bandwidth characteristics and frequency hopping range characteristics of the drone signal; then it is determined whether the number of times the corresponding frames appear meets the occurrence requirements. If so, it means that the signal to be detected also meets the time width characteristics of the drone signal. Since all the characteristics of the signal to be detected are met, it is considered that the signal to be detected is a drone signal, and the drone detection is completed; the processing flow of this application solution is simple and easy to implement, does not require complex processing flow, and the detection time is short; the amount of calculation is small, does not rely on specific hardware, and the implementation cost is low.

[0022] Optionally, the step of performing FFT processing on each frame of time domain data to convert it into frequency domain data includes:

[0023] The time domain data of each frame is summed up according to the preset second division time length to obtain the optimized time domain data;

[0024] Perform FFT processing on the optimized time domain data to obtain the corresponding frequency domain data.

[0025] By adopting the above technical solution, since each frame of time domain data is summed and accumulated, the amount of calculation is reduced, thereby optimizing the overall amount of calculation.

[0026] Optionally, the step of obtaining a denoised signal power spectrum value corresponding to each frame of the frequency domain data according to each of the average channel powers includes:

[0027] Taking the maximum value of the average channel power of each frequency domain channel in the frequency domain data of each frame as the background noise power corresponding to each frequency domain channel in the frequency domain data of each frame;

[0028] Noise is filtered out of each frequency domain channel in each frame of the frequency domain data according to the background noise power to obtain a denoised signal power spectrum value corresponding to each frequency domain channel in each frame of the frequency domain data.

[0029] By adopting the above technical solution, The background noise level is estimated based on the maximum average power of each channel in the frame signal, and the system has the ability to dynamically adjust the background noise power.

[0030] Optionally, the power spectrum value of the denoised signal is , , ;in, is the power spectrum value of the denoised signal of the jth frequency domain channel in the i-th frame frequency domain data, is the background noise power, is the average channel power, is the channel power of the jth frequency domain channel in the i-th frame frequency domain data, is the frequency domain data point contained in the frequency domain channel, is the frequency domain data of the i-th frame The jth frequency domain channel in .

[0031] By adopting the above technical solution, the maximum of the result after power subtraction and 0 in the power spectrum value formula of the denoised signal is taken to ensure that the power spectrum value of the denoised signal is non-negative.

[0032] Optionally, the step of obtaining a detection threshold of the frequency domain data of each frame according to the power spectrum value of the denoised signal includes:

[0033] According to the preset false alarm rate , the proportional factor of the detection threshold is calculated by the inverse error complementary function : ;

[0034] Calculate the denoised power spectrum value in the frequency domain data of each frame The median absolute deviation of ; ;

[0035] According to the scaling factor and the median absolute deviation value , obtain the detection threshold corresponding to the frequency domain data of each frame .

[0036] By adopting the above technical solution and combining it with threshold detection, the reliability of the signal is improved and the detection result is more accurate.

[0037] Optionally, the step of extracting the corresponding candidate channel in each frame of the frequency domain data according to the detection threshold is:

[0038] Comparing the power spectrum value of the denoised signal of each frequency domain channel in each frame of the frequency domain data with the detection threshold;

[0039] The frequency domain channel whose power spectrum value of the denoised signal is greater than the detection threshold is identified as a candidate channel.

[0040] Optionally, the number requirement is:

[0041] ;in, is the broadband characteristic of the UAV signal, is the frequency hopping range characteristic of the drone signal, is the signal sampling rate, is the total number of frames of divided frequency domain data.

[0042] Optionally, the occurrence requirement is: ;in, is the number of occurrences of the corresponding frame frequency domain data whose number of candidate channels meets the number requirement, UAV signal time-width characteristics, The duration of the first division.

[0043] In the second aspect, this application provides a drone signal screening system for drones in the 915MHz frequency band, which adopts the following technical solutions:

[0044] A drone signal screening system for drones operating in the 915MHz frequency band, comprising:

[0045] A signal receiving module, used for receiving a signal to be detected;

[0046] a frequency domain conversion module, configured to divide the signal to be detected into multiple frames of time domain data according to a preset first division time length, and perform FFT processing on each frame of time domain data to convert it into frequency domain data;

[0047] The channel processing module is used to divide the frequency domain data of each frame into multiple frequency domain channels and calculate the channel power of a single frequency domain channel; and to calculate the power of the frequency domain channel within the set range. The average channel power of each of the frequency domain channels in the frequency domain data of the frame; used to obtain a denoised signal power spectrum value corresponding to each frame of the frequency domain data based on each of the average channel powers; used to obtain a detection threshold corresponding to each frame of the frequency domain data based on the denoised signal power spectrum value; used to extract the corresponding candidate channel in each frame of the frequency domain data based on the detection threshold;

[0048] The judging module is used to judge in turn whether there are adjacent candidate channels in the frequency domain data of each frame. If so, the channel processing module calculates the number of adjacent candidate channels in the frequency domain data of each frame. ;

[0049] The judgment module then judges the Whether the number requirement is met; if so, further determine whether the number of occurrences of the frequency domain data in the corresponding frame where the number of candidate channels meets the number requirement meets the occurrence requirement; if so, determine that the signal to be detected is a drone signal of a drone in the 915MHz frequency band.

[0050] By adopting the above technical solution, the received signal of a certain time length is regarded as a frame of data, and FFT (Fast Fourier Transform) is performed to transform the time domain of each frame signal into the frequency domain; then, each frame of frequency domain data is divided into several frequency domain channels, and the average signal power of all frequency domain channels is calculated; then, according to each average channel power, the corresponding denoised signal power spectrum is obtained to further calculate the detection threshold of each frame of frequency domain data; then, the corresponding candidate channels in each frame of frequency domain data are extracted, and whether the candidate channels in each frame are adjacent is determined in turn. If adjacent, the number of adjacent candidate channels in each frame of frequency domain data is calculated. ,like If the number requirements are met, it means that the signal to be detected meets the bandwidth characteristics and frequency hopping range characteristics of the drone signal; then it is determined whether the number of times the corresponding frames appear meets the occurrence requirements. If so, it means that the signal to be detected also meets the time width characteristics of the drone signal. Since all the characteristics of the signal to be detected are met, it is considered that the signal to be detected is a drone signal, and the drone detection is completed; the processing flow of this application solution is simple and easy to implement, does not require complex processing flow, and the detection time is short; the amount of calculation is small, does not rely on specific hardware, and the implementation cost is low.

[0051] In a third aspect, the present application provides a terminal that adopts the following technical solution:

[0052] A terminal, comprising:

[0053] Memory, storing a 915MHz band drone detection program;

[0054] The processor is used to execute the program stored in the memory to implement the steps of the above-mentioned 915MHz frequency band drone detection method.

[0055] In summary, this application has at least the following beneficial effects:

[0056] 1. By combining the characteristic parameters of drones in the 915MHz frequency band to judge the detection signal, it is ensured that only drone signals can be fully matched and other interference signals cannot be matched, thereby reducing the false alarm rate;

[0057] 2. Use based on The maximum average power of each frequency domain channel in the frame frequency domain data is used to estimate the background noise level. It has the ability to dynamically adjust the background noise power. Combined with threshold detection, it improves the reliability of the signal and the accuracy of the detection result.

[0058] 3. Able to adapt to various 915MHz frequency band drone detection, just need to replace different characteristic parameters;

[0059] 4. The processing flow is simple and easy to implement, without the need for complicated processing flow, and the detection time is short;

[0060] 5. Small amount of computation, no reliance on specific hardware, and low implementation cost. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] Figure 1 This is a flowchart of part of the method embodiment of the present application;

[0062] Figure 2 This is another flow chart of the method embodiment of the present application;

[0063] Figure 3 This is the time-frequency diagram of the 915MHz drone signal;

[0064] Figure 4 This is a time domain diagram of a TBS digital transmission signal;

[0065] Figure 5 This is a frequency domain diagram of a TBS digital transmission signal;

[0066] Figure 6 It is the original signal spectrum of the signal to be detected converted from the time domain to the frequency domain;

[0067] Figure 7 It is the spectrum after optimization after 16-point summation;

[0068] Figure 8 It is a schematic diagram of the channel power of a single frequency domain channel;

[0069] Figure 9 It is a comparison diagram of the power spectrum value of the denoised signal and the detection threshold after noise reduction of a single frequency domain channel;

[0070] Figure 10 It is a structural block diagram of an embodiment of the system of the present application.

[0071] Description of reference numerals: 101, signal receiving module; 102, frequency domain conversion module; 103, channel processing module; 104, judgment module. DETAILED DESCRIPTION

[0072] To make the purpose, technical solutions and advantages of the embodiments of the present application clearer, the following will be combined with the appended drawings of the embodiments of the present invention. Figure 1 -Attached Figure 10 The technical solutions in the embodiments of the present invention are clearly and completely described. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of them. All other embodiments derived by persons of ordinary skill in the art based on the embodiments of the present invention without inventive effort are within the scope of protection of the present invention.

[0073] The technical principles of this application are:

[0074] The digital signals transmitted by drones operating in the 915MHz frequency band, such as those equipped with TBS digital transmission modules, exhibit specific time-frequency patterns, whereas non-drone signals do not. Specifically, a single drone-transmitted digital signal has a fixed duration and bandwidth, differing from the center frequency of the next transmitted signal. However, multiple transmitted signals fall within a large frequency range. Therefore, the present invention designs a universal 915MHz drone detection solution based on the time-frequency patterns of digital signals. First, a signal time-frequency model is constructed to analyze the single duration, signal bandwidth and total frequency range of various drone signals in the 915MHz frequency band. Secondly, the received signal of a certain duration is regarded as a frame of data, and the signal is transformed into the frequency domain through FFT processing. Thirdly, each frame of data is divided into several channels, and the average signal power of all channels is calculated. Then, the maximum value of the average power of the most recent frames of signals is taken as the background noise power, and the noise of each frame signal is eliminated based on this value, and the power detection threshold of the signal after noise elimination is calculated. Finally, feature detection is performed on signals with power higher than the detection threshold to determine whether their duration and signal bandwidth meet the characteristics of 915MHz drone signals. When all features meet the requirements, the signal is considered to be a drone signal, and drone detection is completed.

[0075] Based on the above technical principles, the first embodiment of this application discloses a method for detecting drones in the 915MHz frequency band. Figure 1 and Figure 2 As an implementation of the detection method, the detection method may include S101-S117:

[0076] S101, receiving a signal to be detected;

[0077] S102, dividing the signal to be detected into multiple frames of time domain data according to a preset first division time length;

[0078] S103, performing FFT processing on each frame of time domain data to convert it into corresponding frequency domain data;

[0079] S104, dividing each frame of frequency domain data into multiple frequency domain channels;

[0080] S105, calculating the channel power of a single frequency domain channel;

[0081] S106, for the set range Frame frequency domain data, calculating the average channel power of each frequency domain channel in each frame frequency domain data;

[0082] S107, obtaining a denoised signal power spectrum value corresponding to each frame of frequency domain data according to each average channel power;

[0083] S108, obtaining a detection threshold for each frame of frequency domain data according to the power spectrum value of the denoised signal;

[0084] S109, extracting the corresponding candidate channel from each frame of frequency domain data according to the detection threshold;

[0085] S110, determining in sequence whether there are adjacent candidate channels in each frame of frequency domain data;

[0086] S111, if not, discard the corresponding frame frequency domain data;

[0087] S112: If yes, calculate the number of adjacent candidate channels in the corresponding frame frequency domain data of the adjacent candidate channels. ;

[0088] S113, judgment Whether the quantity requirements are met;

[0089] S114, if not, discard the corresponding frame frequency domain data;

[0090] S115, if yes, then determine whether the number of occurrences of the corresponding frame frequency domain data of the candidate channels that meet the number requirement meets the occurrence requirement;

[0091] S116, if yes, then determine that the signal to be detected is a drone signal of a drone in the 915 MHz frequency band;

[0092] S117: If not, the signal to be detected is discarded.

[0093] Specifically, before receiving the signal to be detected, a time-frequency model of the drone signal can be constructed. For drones with a communication frequency in the 915MHz band, the duration and bandwidth of a single transmitted signal are fixed, while the transmission frequency of a single signal varies within a large frequency range. Figure 3 Therefore, the signal time width (denoted as ), signal bandwidth (denoted as ), frequency hopping range (denoted as ) as a feature quantity, pre-collect signal data from various drones equipped with TBS modules and perform feature analysis to construct a signal time-frequency model. During feature analysis, short-time Fourier transform can be used to plot time-frequency plots for time-frequency feature analysis.

[0094] For example, a typical TBS digital signal in time domain and frequency domain can be referred to as Figure 4 and Figure 5 As shown in the figure, after a large amount of data collection and analysis, the three characteristic quantities of drones in the 915 MHz frequency band are: .

[0095] For S102, specifically, the received signal to be detected that may contain a drone signal is recorded as , the first division time length is 8192, for example, then every 8192 points are divided into a frame, and the time domain data is obtained .

[0096] For S103, the specific steps are:

[0097] The time domain data of each frame is summed up according to the preset second division time length to obtain the optimized time domain data;

[0098] Perform FFT processing on the optimized time domain data to obtain frequency domain data.

[0099] Specifically, according to the characteristics of FFT operation, when the input is 8192 points of time domain data, the output is 8192 points of frequency domain data. In order to optimize the subsequent amount of calculation, the data of 8192 points in each frame are summed up. For example, if the second division time length is 16, the sum is calculated every 16 points, and finally a 512-point frequency domain result is obtained for each frame. The frequency domain data of each frame is recorded as .

[0100] The original frequency domain data of the signal to be detected can be referred to Figure 6 , the optimized frequency domain data can be referred to Figure 7 .

[0101] Specifically, for S104, the number of divided channels Depends on the specific drone model and transmission protocol. For example, the TBS 915MHz standard transmission protocol has a signal frequency range of After analyzing a large amount of measured data, it is found that it has exactly 100 frequency domain channels. The starting frequency and ending frequency of each channel are fixed, which just accommodates the bandwidth of a single-segment signal ( ). Therefore, the frequency domain data of each frame signal of the TBS 915MHz standard protocol is converted into Divide the frequency domain channels. For example, the frequency domain data of the i-th frame signal After dividing the frequency domain channels, we get , recorded as Each frequency domain channel contains frequency domain data points.

[0102] For S105, the channel power of a single frequency domain channel . You can refer to Figure 8 .

[0103] For S106, specifically, for Frame frequency domain data, calculate the average channel power of each frequency domain channel in each frame frequency domain data ; ; Generally refers to 20 frames to 50 frames.

[0104] For S107, the specific steps are:

[0105] The maximum value of the average channel power of each frequency domain channel in each frame of frequency domain data is taken as the background noise power corresponding to the channel in each frame of frequency domain data; noise is filtered out for each frequency domain channel in each frame of frequency domain data according to the background noise power, and the denoised signal power spectrum value corresponding to each frequency domain channel in each frame of frequency domain data is obtained.

[0106] Specifically, the background noise power ,according to Filter the noise of each frequency domain channel in each frame of frequency domain data to obtain the corresponding denoised signal power spectrum value The power spectrum value of the denoised signal of the jth frequency domain channel in the i-th frame frequency domain data for:

[0107]

[0108] The maximum of the result after power subtraction and 0 is taken in the formula to ensure that the power spectrum value of the denoised signal is non-negative.

[0109] For S108, the specific steps are:

[0110] According to the preset false alarm rate , the proportional factor of the detection threshold is calculated by the inverse error complementary function : ;

[0111] Calculate the denoised power spectrum value for each frame The median absolute deviation of ; ;

[0112] According to the scale factor and the median absolute deviation , obtain the detection threshold of each frame of frequency domain data .

[0113] For S109, the specific steps are:

[0114] The power spectrum value of the denoised signal of each frequency domain channel in each frame of frequency domain data is compared with the detection threshold; the frequency domain channel whose power spectrum value of the denoised signal is greater than the detection threshold is identified as a candidate channel.

[0115] The comparison chart of the power spectrum value of the denoised signal and the detection threshold after noise reduction of a single frequency domain channel can be referred to Figure 9 .

[0116] After completing the extraction of candidate channels for each frame of frequency domain data, determine whether there are adjacent candidate channels in each frame of frequency domain data. If so, calculate the number of adjacent candidate channels. If the signal sampling rate is ,but Should meet the following requirements:

[0117] .

[0118] According to the above specific example, if is 512, then Should meet .

[0119] When the number of candidate channels meets the number requirement, it means that the frame frequency domain data of the candidate channel meets the UAV signal bandwidth characteristics ( ) and frequency hopping range characteristics ( ). Continue to further analyze the frequency domain data of the subsequent frames. If the number of candidate channels in the signal to be detected meets the requirements, the number of consecutive occurrences of the corresponding frame frequency domain data is satisfy:

[0120] According to the above specific example, if is 8192, then .

[0121] Indicates that the number of candidate channels meets the requirements and the corresponding frame frequency domain data also meets the UAV signal time width characteristics ( ). Since the signal to be detected meets the time-frequency domain characteristics of drone signals, the signal to be detected is considered to be a drone signal.

[0122] Based on the above method embodiment, the second embodiment of the present application discloses a detection system for drones in the 915MHz frequency band. Figure 10 , the detection system may include:

[0123] A signal receiving module 101 is configured to receive a signal to be detected;

[0124] The frequency domain conversion module 102 is used to divide the signal to be detected into multiple frames of time domain data according to a preset first division time length, and perform FFT processing on each frame of time domain data to convert it into frequency domain data;

[0125] The channel processing module 103 is used to divide each frame of frequency domain data into multiple frequency domain channels and calculate the channel power of a single frequency domain channel; and to calculate the power of the frequency domain within a set range. The average channel power of each frequency domain channel in the frame frequency domain data; used to obtain the denoised signal power spectrum value corresponding to each frame of frequency domain data based on the average channel power; used to obtain the detection threshold corresponding to each frame of frequency domain data based on the denoised signal power spectrum value; used to extract the corresponding candidate channel in each frame of frequency domain data based on the detection threshold;

[0126] The judging module 104 is used to judge whether there are adjacent candidate channels in each frame of frequency domain data. If so, the channel processing module calculates the number of adjacent candidate channels in each frame of frequency domain data. ;

[0127] The judgment module 104 judges again Whether the number requirement is met; if so, further determine whether the number of occurrences of the corresponding frame frequency domain data of the candidate channels that meet the number requirement meets the occurrence requirement; if so, determine that the signal to be detected is a drone signal of a drone in the 915MHz frequency band.

[0128] The modules of the 915MHz band drone detection system correspond one to one with the 915MHz band drone detection method, and no further details will be given here.

[0129] A third embodiment of the present application provides a terminal, which may include: a memory and a processor; wherein:

[0130] The memory is used to store the detection program of drones in the 915MHz frequency band;

[0131] The processor is used to execute the program stored in the memory to implement the steps of the above-mentioned 915MHz frequency band drone detection method.

[0132] The memory may be communicatively connected to the processor via a communication bus, and the communication bus may be an address bus, a data bus, a control bus, or the like.

[0133] In addition, the memory may include a random access memory (RAM) and may also include a non-volatile memory (NVM), such as at least one disk storage.

[0134] The processor can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc.

[0135] The above are all preferred embodiments of the present application and are not intended to limit the scope of protection of the present application. Unless otherwise specified, any feature disclosed in this specification (including the abstract and drawings) may be replaced by other equivalent or similar features. In other words, unless otherwise specified, each feature is merely an example of a series of equivalent or similar features.

Claims

1. A method for detecting drones in the 915MHz frequency band, characterized in that: include: receiving a signal to be detected; Dividing the signal to be detected into multiple frames of time domain data according to a preset first division time length; Perform fast Fourier transform on each frame of time domain data to convert it into frequency domain data; Divide each frame of frequency domain data into multiple frequency domain channels; Calculating the channel power of a single frequency domain channel; Calculate the set range an average channel power of each of the frequency domain channels in the frequency domain data of the frame; Obtaining a denoised signal power spectrum value corresponding to each frame of the frequency domain data according to each of the average channel powers, wherein the denoised signal power spectrum value is a non-negative value; Obtaining a detection threshold for each frame of the frequency domain data according to the power spectrum value of the denoised signal; extracting corresponding candidate channels from each frame of the frequency domain data according to the detection threshold; determining in sequence whether there is an adjacent candidate channel in each frame of the frequency domain data; If so, calculate the number of adjacent candidate channels in the frequency domain data of the corresponding frame of the adjacent candidate channels. ; Judge the Whether the quantity requirements are met; If so, determining whether the number of occurrences of the frequency domain data in the corresponding frame where the number of candidate channels meets the number requirement meets the occurrence requirement; If so, it is determined that the signal to be detected is a drone signal of a drone in the 915MHz frequency band; The step of obtaining the denoised signal power spectrum value corresponding to each frame of the frequency domain data according to each of the average channel powers comprises: Taking the maximum value of the average channel power of each frequency domain channel in the frequency domain data of each frame as the background noise power corresponding to each frequency domain channel in the frequency domain data of each frame; The number requirements are: ;in, is the broadband characteristic of the UAV signal, is the frequency hopping range characteristic of the drone signal, is the signal sampling rate, n is the total number of frames of divided frequency domain data; The requirements for appearance are: Wherein, Cnt is the number of occurrences of the corresponding frame frequency domain data whose number of candidate channels meets the number requirement, is the time width feature of the UAV signal, and c is the duration of the first division.

2. The method for detecting drones in the 915MHz frequency band according to claim 1, characterized in that: The step of performing FFT processing on each frame of time domain data to convert it into frequency domain data includes: The time domain data of each frame is summed up according to the preset second division time length to obtain the optimized time domain data; Perform fast Fourier transform on the optimized time domain data to obtain the corresponding frequency domain data.

3. The method for detecting drones in the 915MHz frequency band according to claim 1, characterized in that: The step of obtaining the denoised signal power spectrum value corresponding to each frame of the frequency domain data according to each of the average channel powers further includes: Noise is filtered out of each frequency domain channel in each frame of the frequency domain data according to the background noise power to obtain a denoised signal power spectrum value corresponding to each frequency domain channel in each frame of the frequency domain data.

4. The method for detecting drones in the 915MHz frequency band according to claim 3, characterized in that: The power spectrum of the denoised signal is , , ;in, is the power spectrum value of the denoised signal of the jth frequency domain channel in the i-th frame frequency domain data, is the background noise power, is the average channel power, is the channel power of the jth frequency domain channel in the i-th frame frequency domain data, is the frequency domain data point contained in the frequency domain channel, is the frequency domain data of the i-th frame The jth frequency domain channel in .

5. The method for detecting drones in the 915MHz frequency band according to claim 4, characterized in that: The step of obtaining a detection threshold of the frequency domain data of each frame according to the power spectrum value of the denoised signal comprises: According to the preset false alarm rate , the proportional factor of the detection threshold is calculated by the inverse error complementary function ; Calculate the median absolute deviation of the denoised power spectrum values ​​in the frequency domain data of each frame ; According to the scale factor and the median absolute deviation value, the detection threshold corresponding to each frame of the frequency domain data is obtained. .

6. The method for detecting drones in the 915MHz frequency band according to claim 1, characterized in that: The step of extracting the corresponding candidate channel in each frame of the frequency domain data according to the detection threshold is: Comparing the power spectrum value of the denoised signal of each frequency domain channel in each frame of the frequency domain data with the detection threshold; The frequency domain channel whose power spectrum value of the denoised signal is greater than the detection threshold is identified as a candidate channel.

7. A 915MHz band drone detection system, characterized in that: The method for detecting a drone in the 915 MHz frequency band according to any one of claims 1 to 6 comprises: A signal receiving module (101), configured to receive a signal to be detected; A frequency domain conversion module (102) is used to divide the signal to be detected into multiple frames of time domain data according to a preset first division time length, and perform fast Fourier transform processing on each frame of time domain data to convert it into frequency domain data; The channel processing module (103) is used to divide the frequency domain data of each frame into multiple frequency domain channels and calculate the channel power of a single frequency domain channel; and to calculate the power of the frequency domain channel within a set range. The average channel power of each of the frequency domain channels in the frequency domain data of the frame; used to obtain a denoised signal power spectrum value corresponding to each frame of the frequency domain data based on each of the average channel powers; used to obtain a detection threshold corresponding to each frame of the frequency domain data based on the denoised signal power spectrum value; used to extract the corresponding candidate channel in each frame of the frequency domain data based on the detection threshold; The judging module (104) is used to judge in turn whether there are adjacent candidate channels in each frame of the frequency domain data. If so, the channel processing module calculates the number of adjacent candidate channels in each frame of the frequency domain data. ; The judgment module (104) then judges the Whether the number requirement is met; if so, further determine whether the number of occurrences of the frequency domain data in the corresponding frame where the number of candidate channels meets the number requirement meets the occurrence requirement; if so, determine that the signal to be detected is a drone signal of a drone in the 915MHz frequency band.

8. A terminal, characterized in that: include: Memory, storing a 915MHz band drone detection program; A processor, configured to execute the program stored on the memory to implement the steps of the 915 MHz frequency band drone detection method according to any one of claims 1 to 6.

Images