UAV detection method, system and related equipment based on software radio
Through a software radio-based method, signal preprocessing and clustering technology is used to realize drone detection, solving the problems of complexity and cost of existing systems, and improving the flexibility and general utility of the detection system.
Patent Information
- Application Number
- CN202210536510.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2022-05-17
- Publication Date
- 2024-11-22
- Estimated Expiration
- 2042-05-17
AI Technical Summary
The existing electromagnetic wave-based drone detection system is complex and has high implementation costs, which is difficult to meet the detection needs of individual users and is not conducive to the expansion of subsequent functions.
Using a software radio-based method, the radio signals are acquired for preprocessing, intra-segment signal clustering, inter-segment signal clustering and inter-class signal clustering, and finally classified and discriminated to realize drone detection. The method includes signal segmentation, Fourier transform, dynamic denoising, binarization and clustering processing, reducing the cost and implementation complexity of the hardware architecture.
It improves the flexibility and universality of the drone detection system, reduces hardware costs, and realizes flexible drone signal recognition and detection.
Smart Images

Figure CN115079150B_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the field of radio technology application, and in particular relates to a software radio-based unmanned aerial vehicle detection method, system and related equipment. Background Art
[0002] In recent years, small civilian drones have developed rapidly and are used in all walks of life, such as aerial photography, flight performances, etc. However, due to the lack of management methods and relevant regulations, the safety issues brought by small civilian drones have become increasingly prominent, especially for drones assembled by individuals without identification information. The phenomenon of illegal flying is particularly serious. For example, the recent frequent illegal intrusions into airports and drone sneak shots have seriously endangered the safety of the public and personal privacy.
[0003] In this context, the corresponding countermeasures against the phenomenon of illegal drone flights are particularly important, and the premise of countermeasures is to conduct identification and detection first. Among various detection methods for drones, the method of using electromagnetic waves is widely used because of its all-weather performance that is not affected by weather and the real-time feedback of results. However, from the existing research schemes, this method also has complex detection methods. For example, in order to improve the recognition accuracy of electromagnetic waves, algorithms such as machine learning are used, the electromagnetic signal recognition of different drones is not flexible enough, and the cost of related hardware is high. It is not conducive to the expansion of subsequent functions other than detection. Individual users want to use electromagnetic waves for drone detection but are always discouraged by the above problems. Summary of the invention
[0004] The embodiments of the present invention provide a software radio-based drone detection method, system and related equipment, aiming to solve the problems of complexity and high implementation cost of traditional electromagnetic wave-based drone detection systems.
[0005] In a first aspect, an embodiment of the present invention provides a method for detecting a drone based on software defined radio, the method comprising the following steps:
[0006] S1. Acquire a radio signal and preprocess the radio signal to obtain a preprocessed signal, wherein the parameters of the radio signal include a power value, a frequency value, a time value, a bandwidth and a pulse width, the frequency value includes an upper frequency limit and a lower frequency limit, and the time value includes an arrival time and an end time;
[0007] S2, performing intra-segment signal clustering on the preprocessed signal to obtain a first cluster signal;
[0008] S3, performing inter-segment signal clustering on the first cluster signal to obtain a second cluster signal;
[0009] S4, performing inter-class signal clustering on the second clustering signal to obtain a third clustering signal;
[0010] S5. Classify and identify the third cluster signal to obtain a drone detection result.
[0011] Furthermore, the step of acquiring a radio signal and preprocessing the radio signal to obtain a preprocessed signal comprises the following sub-steps:
[0012] S11, segmenting the radio signal according to a sub-signal length M to obtain multiple sub-signals, wherein the sub-signals include at least a continuous first sub-signal and a second sub-signal, and performing Fourier transform processing on each sub-signal;
[0013] S12, dynamically denoising the sub-signal after Fourier transformation, wherein the point power of each point of the sub-signal within its length M is y, and the maximum power y of each sub-signal in its frequency domain is calculated. max , and calculate the power mean y according to the maximum power mean ;
[0014] S13, judge y max / y mean The value of is less than the preset threshold value, and the dynamic threshold value is set, where:
[0015] If y max / y mean If the value of is less than the preset threshold value, the value of the dynamic threshold value is set to y max +1;
[0016] If y max / y mean If the value of is equal to or greater than the preset threshold value, the value of the dynamic threshold value is set to y mean A value obtained by multiplying the preset threshold value;
[0017] S14. Binarize the power value of the sub-signal according to the dynamic threshold value to obtain a binarized sub-signal.
[0018] Furthermore, the signal overlap between the consecutive sub-signals is 50% each, wherein, for the sub-signals of length M, the point power y of each point has the maximum power y max , the maximum power y max The following relationship (1) is satisfied:
[0019] y max =max{y 2 (k)},k∈(0,1,…,M-1) (1);
[0020] For the point power, define the point powers of the first sub-signal and the second sub-signal as y1 and y2 respectively, calculate y1 / y2 for the sub-signals within a range of length M, and obtain a new point power y with a length of M. new , wherein, if y1 / y2 is greater than or equal to the preset power threshold value, the new point power mark starting from y1 will be set to clear_flag until y1 / y2 is greater than the preset power threshold value, and the new point power satisfies the following relationship (2):
[0021]
[0022] The power mean y mean The value of is equal to y(k) new After accumulation, divide by y(k) new The number of non-zero values in ;
[0023] The preset threshold value is defined as sel_th, and the dynamic threshold value is defined as noise_th. Then, the dynamic threshold value noise_th satisfies the following relation (3):
[0024]
[0025] According to the dynamic threshold value, each point power y of the sub-signal is binarized to obtain the binarized sub-signal, wherein y is set to 1 when it is greater than the dynamic threshold value, and is set to 0 when it is equal to or less than the dynamic threshold value.
[0026] Furthermore, in step S2, the step of performing intra-segment signal clustering on the preprocessed signal to obtain a first cluster signal includes the following sub-steps:
[0027] S21, obtaining the frequency value of the point power of the first non-zero point in the binary sub-signal, and recording the frequency value in a first cluster storage unit, wherein the first cluster storage unit is a column structure, and the frequency value is stored in the first column of the first cluster storage unit, and at the same time, the frequency value of the point power of the first non-zero point in the binary sub-signal is used as a first reference value;
[0028] S22, obtaining the frequency value of the point power of the next non-zero point in the binarized sub-signal that is not stored in the first cluster storage unit, and performing a difference calculation between the frequency value and the first reference value, wherein:
[0029] If the difference between the frequency value and the first reference value is less than or equal to a first preset frequency clustering threshold, assigning the frequency upper limit in the last non-empty column in the first clustering storage unit with the value of the frequency upper limit of the frequency value;
[0030] If the difference between the frequency value and the first reference value is greater than the first preset frequency clustering threshold, assigning the value of the frequency value to the first reference value, and at the same time, recording the sub-signal corresponding to the frequency value into the first empty column of the first cluster storage unit;
[0031] S23, repeating step S22 until the difference comparison between all the point powers in the binarized sub-signal and the first reference value is completed, and then the first cluster storage unit is output as the first cluster signal.
[0032] Furthermore, in step S3, the step of performing inter-segment signal clustering on the first cluster signal to obtain a second cluster signal includes the following sub-steps:
[0033] S31, taking the frequency value and time value of the first column in the first cluster signal as the second reference value, and recording the sub-signal corresponding to the first column in the first cluster signal into a second cluster storage unit, the second cluster storage unit being a column structure, wherein the second column in the first cluster signal is taken as the current column of the inter-segment signal clustering;
[0034] S32, comparing the frequency value and the time value of the sub-signal in the current column with the second reference value, wherein:
[0035] If the difference between the frequency value and the second reference value is less than the second preset clustering threshold, and the difference between the time value and the second reference value is less than the preset time clustering threshold, then continue to determine whether the upper frequency limit in the frequency value is greater than the second reference value, wherein if so, assign the value of the upper frequency limit to the second reference value; if not, keep the second reference value unchanged;
[0036] If the difference between the frequency value and the second reference value is greater than or equal to the second preset clustering threshold, or the difference between the time value and the second reference value is greater than or equal to the preset time clustering threshold, storing the sub-signal corresponding to the current column into the first empty column in the second clustering storage unit;
[0037] After the difference between the frequency value and the time value of the current column and the second reference value is compared, the next column of the current column is used as the new current column;
[0038] S33, repeating step S32 until all the sub-signals in the first cluster signal are compared with the second reference value, and then outputting all the sub-signals in the second cluster storage unit as the second cluster signal.
[0039] Furthermore, in step S4, the step of performing inter-class signal clustering on the second cluster signal to obtain a third cluster signal includes the following sub-steps:
[0040] S41, taking the pulse width of the first column in the second cluster signal as a third reference value, and recording the sub-signal corresponding to the first column in the second cluster signal into a third cluster storage unit, wherein the third cluster storage unit is a column structure, wherein the second column in the second cluster signal is used as a current column for clustering inter-class signals;
[0041] S42, comparing the pulse width of the sub-signal in the current column with the third reference value, wherein:
[0042] If the difference between the pulse width and the third reference value is smaller than a preset pulse width clustering threshold, storing the sub-signal corresponding to the current column into a first empty column of the third clustering storage unit;
[0043] If the difference between the pulse width and the third reference value is greater than or equal to the preset pulse width clustering threshold, assigning the value of the pulse width to the third reference value;
[0044] After the difference between the pulse width of the current column and the third reference value is compared, the next column of the current column is used as the new current column;
[0045] S43, repeating step S42 until all the sub-signals in the second cluster signal are compared with the third reference value, and outputting all the sub-signals in the third cluster storage unit as the third cluster signal.
[0046] Furthermore, in step S5, the step of classifying and distinguishing the third cluster signal to obtain the drone detection result includes the following sub-steps:
[0047] S51, taking out all the sub-signals in the third clustering signal to obtain a one-dimensional array of S columns, subtracting the parameter values of the sub-signals in the first column of the one-dimensional array from the parameter values corresponding to all columns except the first column, and at the same time, subtracting the parameter values of the sub-signals in the second column of the one-dimensional array from the parameter values corresponding to all columns except the first column and the second column to obtain (S-1)*S / 2 feature judgment difference values, and grouping the feature judgment difference values according to a preset feature judgment threshold to obtain N difference value groups;
[0048] S52, the feature judgment difference corresponding to the group containing the largest number of feature judgment difference values in the difference grouping is recorded as the pulse interval time, and the pulse width and the pulse interval time of every two adjacent sub-signals in the one-dimensional array are respectively calculated with the preset drone judgment signal, wherein:
[0049] If the difference between the pulse width and the preset drone judgment signal is less than or equal to the preset feature judgment difference, or the difference between the pulse interval time and the preset drone judgment signal is less than or equal to the preset feature judgment difference, it is determined that there is a drone signal between the corresponding two adjacent sub-signals;
[0050] If the difference between the pulse width or the inter-pulse interval time of the sub-signal and the preset drone judgment signal is greater than the preset feature judgment difference, it is determined that there is no drone signal between the corresponding two adjacent sub-signals.
[0051] In a second aspect, an embodiment of the present invention further provides a drone detection system based on software radio, comprising:
[0052] A signal preprocessing module, used to acquire a radio signal and preprocess the radio signal to obtain a preprocessed signal, wherein the parameters of the radio signal include a power value, a frequency value, a time value, a bandwidth and a pulse width, wherein the frequency value also includes an upper frequency limit and a lower frequency limit, and the time value also includes an arrival time and an end time;
[0053] An intra-segment clustering module, used for performing intra-segment signal clustering on the preprocessed signal to obtain a first clustering signal;
[0054] An inter-segment clustering module, used for performing inter-segment signal clustering on the first clustering signal to obtain a second clustering signal;
[0055] An inter-class clustering module, used for performing inter-class signal clustering on the second clustering signal to obtain a third clustering signal;
[0056] The cluster detection module is used to classify and identify the third cluster signal to obtain a drone detection result.
[0057] In a third aspect, an embodiment of the present invention further provides a computer device, comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the steps in the software radio-based drone detection method as described in any one of the above embodiments are implemented.
[0058] In a fourth aspect, an embodiment of the present invention further provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps in the software radio-based drone detection method as described in any one of the above embodiments are implemented.
[0059] The beneficial effects achieved by the present invention are that the software radio-based method is used to realize drone detection, and three clustering methods for different radio signal parameters are combined to distinguish drone signals. While improving the flexibility of the detection system, the cost of the hardware architecture is also reduced, and the versatility of the drone detection system can be improved. BRIEF DESCRIPTION OF THE DRAWINGS
[0060] Figure 1 It is a flowchart of the steps of the software radio-based drone detection method provided by an embodiment of the present invention;
[0061] Figure 2 It is a flowchart of the sub-steps of step S1 in the software radio-based drone detection method provided in an embodiment of the present invention;
[0062] Figure 3 It is a flowchart of the sub-steps of step S2 in the software radio-based drone detection method provided in an embodiment of the present invention;
[0063] Figure 4 It is a flowchart of the sub-steps of step S3 in the software radio-based drone detection method provided in an embodiment of the present invention;
[0064] Figure 5 It is a flowchart of the sub-steps of step S4 in the software radio-based drone detection method provided in an embodiment of the present invention;
[0065] Figure 6 It is a flowchart of the sub-steps of step S5 in the software radio-based drone detection method provided in an embodiment of the present invention;
[0066] Figure 7 is a structural schematic diagram of a software radio-based drone detection system 200 provided in an embodiment of the present invention;
[0067] Figure 8 It is a schematic diagram of the structure of a computer device provided in an embodiment of the present invention. DETAILED DESCRIPTION
[0068] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.
[0069] Please refer to Figure 1 , Figure 1 The following is a flowchart of a method for detecting a drone based on software defined radio according to an embodiment of the present invention, comprising the following steps:
[0070] S1. Acquire a radio signal and preprocess the radio signal to obtain a preprocessed signal. The parameters of the radio signal include a power value, a frequency value, a time value, a bandwidth and a pulse width. The frequency value includes an upper frequency limit and a lower frequency limit. The time value includes an arrival time and an end time.
[0071] Please refer to Figure 2 , Figure 2 : is a flowchart of sub-steps of step S1 in the software radio-based drone detection method provided by an embodiment of the present invention, and step S1 also includes the following sub-steps:
[0072] S11. Segment the radio signal according to a sub-signal length M to obtain multiple sub-signals, wherein the sub-signals at least include a continuous first sub-signal and a second sub-signal, and perform Fourier transform processing on each sub-signal.
[0073] More specifically, in an embodiment of the present invention, the signal overlap between consecutive sub-signals is 50% each. For example, the received radio signal is segmented into M points, and a fifty percent signal overlap is set between consecutive sub-signals. The specific segmentation is as follows: [0, 1, ...M-1], [M / 2, M / 2+1, ...3M / 2-1], ...
[0074] The above segmented sub-signals are subjected to Fourier transform, and the dynamic denoising process of step S12 is performed on each segment starting from the first segment of the sub-signal.
[0075] S12, dynamically denoising the sub-signal after Fourier transformation, wherein the point power of each point of the sub-signal within its length M is y, and the maximum power y of each sub-signal in its frequency domain is calculated. max , and calculate the power mean y according to the maximum power mean .
[0076] More specifically, for the point power y of each point in the sub-signal with a length of M, the point power y has the maximum power y max , the maximum power y max The following relationship (1) is satisfied:
[0077] y max =max{y 2 (k)},k∈(0,1,…,M-1) (1);
[0078] For the point power, define the point powers of the first sub-signal and the second sub-signal as y1 and y2 respectively, calculate y1 / y2 for the sub-signals within a range of length M, and obtain a new point power y with a length of M.new , wherein, if y1 / y2 is greater than or equal to the preset power threshold value, the new point power mark starting from y1 will be set to clear_flag until y1 / y2 is greater than the preset power threshold value, and the new point power satisfies the following relationship (2):
[0079]
[0080] The power mean y mean The value of is equal to y(k) new After accumulation, divide by y(k) new The number of non-zero values in .
[0081] S13, judge y max / y mean The value of is less than the preset threshold value, and the dynamic threshold value is set, where:
[0082] If y max / y mean If the value of is less than the preset threshold value, the value of the dynamic threshold value is set to y max +1;
[0083] If y max / y mean If the value of is equal to or greater than the preset threshold value, the value of the dynamic threshold value is set to y mean The value obtained by multiplying the preset threshold value.
[0084] S14. Binarize the power value of the sub-signal according to the dynamic threshold value to obtain a binarized sub-signal.
[0085] The preset threshold value is defined as sel_th, and the dynamic threshold value is defined as noise_th. Then, the dynamic threshold value noise_th satisfies the following relation (3):
[0086]
[0087] According to the dynamic threshold value, each point power y of the sub-signal is binarized to obtain the binarized sub-signal, wherein y is set to 1 when it is greater than the dynamic threshold value, and is set to 0 when it is equal to or less than the dynamic threshold value.
[0088] S2. Perform intra-segment signal clustering on the preprocessed signal to obtain a first cluster signal.
[0089] Please refer to Figure 3 , Figure 3: is a flowchart of sub-steps of step S2 in the software radio-based drone detection method provided by an embodiment of the present invention, and step S2 also includes the following sub-steps:
[0090] S21. Obtain the frequency value of the point power of the first non-zero point in the binary sub-signal, and record the frequency value in a first cluster storage unit, wherein the first cluster storage unit is a column structure, and the frequency value is stored in the first column of the first cluster storage unit. At the same time, the frequency value of the point power of the first non-zero point in the binary sub-signal is used as a first reference value.
[0091] Specifically, the sub-power of the binary sub-signal is assigned a value of 0 or 1 in step S14. In this step, the frequency value of the sub-power of the first non-zero point is taken as the first reference value, and the other sub-powers are compared with the first reference value to achieve clustering of signals within the segment.
[0092] S22, obtaining the frequency value of the point power of the next non-zero point in the binarized sub-signal that is not stored in the first cluster storage unit, and performing a difference calculation between the frequency value and the first reference value, wherein:
[0093] If the difference between the frequency value and the first reference value is less than or equal to the first preset frequency clustering threshold, the frequency upper limit value of the frequency value is assigned to the frequency upper limit in the last non-empty column in the first clustering storage unit. Specifically, in an embodiment of the present invention, the corresponding parameter for comparison with the first reference value is related to the first reference value, that is, when performing intra-segment signal clustering, the parameter for mutual comparison between the sub-signals is the frequency value.
[0094] If the difference between the frequency value and the first reference value is greater than the first preset frequency clustering threshold, the value of the frequency value is assigned to the first reference value, and at the same time, the sub-signal corresponding to the frequency value is recorded in the first empty column of the first clustering storage unit. In the embodiment of the present invention, if the first reference value encounters a value of the frequency value as the comparison object that is greater than the first reference value during comparison, it is necessary to update the value of the first reference value, and assign the frequency value as the comparison object to the first reference value, so as to facilitate the subsequent comparison of the sub-power between the consecutive sub-signals.
[0095] S23, repeating step S22 until the difference comparison between all the point powers in the binarized sub-signal and the first reference value is completed, and then the first cluster storage unit is output as the first cluster signal.
[0096] Generally, for the point power, the corresponding time value is related to its segment number (the sequence number in the length M), the sub-signal length, and the sampling frequency. In the embodiment of the present invention, Fs represents the sampling frequency, and the time value satisfies:
[0097] Time value = (segment number - 1) * (M / 2Fs);
[0098] In addition, after the above step S23 is completed, the processing step of step S2 should continue to be performed on the subsequent signal of the sub-signal to achieve continuous processing of signal clustering.
[0099] S3. Perform inter-segment signal clustering on the first cluster signal to obtain a second cluster signal.
[0100] Please refer to Figure 4 , Figure 4 : is a flowchart of sub-steps of step S3 in the software radio-based drone detection method provided in an embodiment of the present invention. Step S3 also includes the following sub-steps:
[0101] S31. Use the frequency value and time value of the first column in the first cluster signal as the second reference value, and record the sub-signal corresponding to the first column in the first cluster signal into a second cluster storage unit, wherein the second cluster storage unit is a column structure, wherein the second column in the first cluster signal is used as the current column for inter-segment signal clustering.
[0102] Specifically, in an embodiment of the present invention, the parameters used for the benchmark values for comparison in steps S2, S3, and S4 are different. In step S31, the frequency value and the time value of the sub-signal are used as the second benchmark value for inter-segment signal clustering. At the same time, the first column of data in the first clustering signal is used as the initial second benchmark value, and the second column of data is used as the comparison object to perform continuous inter-segment signal clustering.
[0103] S32, comparing the frequency value and the time value of the sub-signal in the current column with the second reference value, wherein:
[0104] If the difference between the frequency value and the second reference value is less than the second preset clustering threshold, and the difference between the time value and the second reference value is less than the preset time clustering threshold, then continue to determine whether the upper frequency limit in the frequency value is greater than the second reference value, wherein if so, the value of the upper frequency limit is assigned to the second reference value; if not, the second reference value is kept unchanged. In the current case, as a condition for changing the second reference value, it is required that the difference between the time value and the second reference value is less than the preset time clustering threshold, and the upper frequency limit is greater than the second reference value.
[0105] If the difference between the frequency value and the second reference value is greater than or equal to the second preset clustering threshold, or the difference between the time value and the second reference value is greater than or equal to the preset time clustering threshold, storing the sub-signal corresponding to the current column into the first empty column in the second clustering storage unit;
[0106] After the difference between the frequency value and the time value of the current column and the second reference value is compared, the next column of the current column is used as the new current column.
[0107] S33, repeating step S32 until all the sub-signals in the first cluster signal are compared with the second reference value, and then outputting all the sub-signals in the second cluster storage unit as the second cluster signal.
[0108] S4. Perform inter-cluster signal clustering on the second cluster signal to obtain a third cluster signal.
[0109] Please refer to Figure 5 , Figure 5 : is a flowchart of sub-steps of step S4 in the software radio-based drone detection method provided in an embodiment of the present invention. Step S4 also includes the following sub-steps:
[0110] S41. Use the pulse width of the first column in the second cluster signal as the third reference value, and record the sub-signal corresponding to the first column in the second cluster signal into a third cluster storage unit, wherein the third cluster storage unit is a column structure, wherein the second column in the second cluster signal is used as the current for clustering inter-class signals.
[0111] Specifically, in this step, the pulse width of the sub-signal is used as the third benchmark value for inter-class signal clustering. At the same time, the first column of data in the second clustering signal is used as the initial third benchmark value, and the second column of data is used as a comparison object to perform continuous inter-class signal clustering.
[0112] S42, comparing the pulse width of the sub-signal in the current column with the third reference value, wherein:
[0113] If the difference between the pulse width and the third reference value is smaller than a preset pulse width clustering threshold, storing the sub-signal corresponding to the current column into a first empty column of the third clustering storage unit;
[0114] If the difference between the pulse width and the third reference value is greater than or equal to the preset pulse width clustering threshold, assigning the value of the pulse width to the third reference value;
[0115] After the difference between the pulse width of the current column and the third reference value is compared, the next column of the current column is used as the new current column;
[0116] S43, repeating step S42 until all the sub-signals in the second cluster signal are compared with the third reference value, and outputting all the sub-signals in the third cluster storage unit as the third cluster signal.
[0117] S5. Classify and identify the third cluster signal to obtain a drone detection result.
[0118] Please refer to Figure 6 , Figure 6 : is a flowchart of sub-steps of step S5 in the software radio-based drone detection method provided by an embodiment of the present invention, and step S5 also includes the following sub-steps:
[0119] S51. Take out all the sub-signals in the third clustering signal to obtain a one-dimensional array of S columns, and make a difference between the parameter values of the sub-signals in the first column of the one-dimensional array and the corresponding parameter values of all columns except the first column. At the same time, make a difference between the parameter values of the sub-signals in the second column of the one-dimensional array and the corresponding parameter values of all columns except the first column and the second column to obtain (S-1)*S / 2 feature judgment difference values, and group the feature judgment difference values according to the preset feature judgment threshold to obtain N difference value groups.
[0120] Exemplarily, when the length is M, the radio signal is divided, and finally the third cluster signal has 4 rows of data, each row of data includes 10 data including various parameters. In step S51, all columns except the first column are subtracted from the first column to form 9 feature judgment difference values, and all columns except the first and second columns are subtracted from the second column to form 8 feature judgment difference values, and so on, generating a total of 45 feature judgment difference values.
[0121] S52. The feature judgment difference corresponding to the group containing the largest number of feature judgment differences in the difference grouping is recorded as the inter-pulse interval time, and the pulse width and the inter-pulse interval time of every two adjacent sub-signals in the one-dimensional array are respectively calculated with the difference of the preset UAV judgment signal.
[0122] In the embodiment of the present invention, the feature judgment threshold is set to 0.1ms, and the 45 feature judgment differences are grouped. Considering that the signals to be identified in each group are the mainstream, and the number of some burst interference signals is small, the feature judgment difference corresponding to the group with the largest number is used as the inter-pulse interval time in the signal identification parameter, wherein:
[0123] If the difference between the pulse width and the preset drone judgment signal is less than or equal to the preset feature judgment difference, or the difference between the pulse interval time and the preset drone judgment signal is less than or equal to the preset feature judgment difference, it is determined that there is a drone signal between the corresponding two adjacent sub-signals. In an embodiment of the present invention, the preset drone judgment signal is a plurality of existing drone signal parameters, and the preset drone judgment signal is equivalent to a database of drone signals, which is used to compare the pulse width of the sub-signal obtained in the embodiment of the present invention with the preset feature judgment difference to determine whether the current sub-signal is the same as the existing drone signal.
[0124] If the difference between the pulse width or the inter-pulse interval time of the sub-signal and the preset drone judgment signal is greater than the preset feature judgment difference, it is determined that there is no drone signal between the corresponding two adjacent sub-signals.
[0125] The beneficial effects achieved by the present invention are that the software radio-based method is used to realize drone detection, and three clustering methods for different radio signal parameters are combined to distinguish drone signals. While improving the flexibility of the detection system, the cost of the hardware architecture is also reduced, and the versatility of the drone detection system can be improved.
[0126] The embodiment of the present invention also provides a drone detection system based on software radio, please refer to Figure 7 , Figure 7 is a schematic diagram of the structure of a UAV detection system 200 based on software radio provided in an embodiment of the present invention, wherein the UAV detection system 200 includes:
[0127] The signal preprocessing module 201 is used to acquire a radio signal and preprocess the radio signal to obtain a preprocessed signal. The parameters of the radio signal include a power value, a frequency value, a time value, a bandwidth and a pulse width. The frequency value also includes an upper frequency limit and a lower frequency limit. The time value also includes an arrival time and an end time.
[0128] The intra-segment clustering module 202 is used to perform intra-segment signal clustering on the pre-processed signal to obtain a first cluster signal;
[0129] An inter-segment clustering module 203, configured to perform inter-segment signal clustering on the first clustering signal to obtain a second clustering signal;
[0130] An inter-class clustering module 204, configured to perform inter-class signal clustering on the second clustering signal to obtain a third clustering signal;
[0131] The cluster detection module 205 is used to classify and identify the third cluster signal to obtain a drone detection result.
[0132] The software radio-based drone detection system 200 can implement the steps in the software radio-based drone detection method in the above embodiment, and can achieve the same technical effects. Please refer to the description in the above embodiment and will not be repeated here.
[0133] The embodiment of the present invention also provides a computer device, please refer to Figure 8 , Figure 8 300 is a schematic diagram of the structure of a computer device provided in an embodiment of the present invention. The computer device 300 includes: a memory 302, a processor 301, and a computer program stored in the memory 302 and executable on the processor 301.
[0134] The processor 301 calls the computer program stored in the memory 302 to execute the steps of the drone detection method based on software radio provided in the embodiment of the present invention. Figure 1 , including:
[0135] S1. Acquire a radio signal and preprocess the radio signal to obtain a preprocessed signal. The parameters of the radio signal include a power value, a frequency value, a time value, a bandwidth and a pulse width. The frequency value includes an upper frequency limit and a lower frequency limit. The time value includes an arrival time and an end time.
[0136] Furthermore, the step of acquiring a radio signal and preprocessing the radio signal to obtain a preprocessed signal comprises the following sub-steps:
[0137] S11, segmenting the radio signal according to a sub-signal length M to obtain multiple sub-signals, wherein the sub-signals include at least a continuous first sub-signal and a second sub-signal, and performing Fourier transform processing on each sub-signal;
[0138] S12, dynamically denoising the sub-signal after Fourier transformation, wherein the point power of each point of the sub-signal within its length M is y, and the maximum power y of each sub-signal in its frequency domain is calculated. max , and calculate the power mean y according to the maximum power mean ;
[0139] S13, judge y max / y mean The value of is less than the preset threshold value, and the dynamic threshold value is set, where:
[0140] If y max / y mean If the value of is less than the preset threshold value, the value of the dynamic threshold value is set to y max +1;
[0141] If y max / y mean If the value of is equal to or greater than the preset threshold value, the value of the dynamic threshold value is set to y mean A value obtained by multiplying the preset threshold value;
[0142] S14. Binarize the power value of the sub-signal according to the dynamic threshold value to obtain a binarized sub-signal.
[0143] Furthermore, the signal overlap between the consecutive sub-signals is 50% each, wherein, for the point power y of each point in the sub-signal with a length of M, the point power y has the maximum power y max , the maximum power y max The following relationship (1) is satisfied:
[0144] y max =max{y 2 (k)},k∈(0,1,…,M-1) (1);
[0145] For the point power, define the point powers of the first sub-signal and the second sub-signal as y1 and y2 respectively, calculate y1 / y2 for the sub-signals within a range of length M, and obtain a new point power y with a length of M. new , wherein, if y1 / y2 is greater than or equal to the preset power threshold value, the new point power mark starting from y1 is set to clear_flag until y1 / y2 is greater than the preset power threshold value, and the new point power value satisfies the following relationship (2):
[0146]
[0147] The power mean y mean The value of is equal to y(k) new After accumulation, divide by y(k) new The number of non-zero values in ;
[0148] The preset threshold value is defined as sel_th, and the dynamic threshold value is defined as noise_th. Then, the dynamic threshold value noise_th satisfies the following relation (3):
[0149]
[0150] According to the dynamic threshold value, each point power y of the sub-signal is binarized to obtain the binarized sub-signal, wherein y is set to 1 when it is greater than the dynamic threshold value, and is set to 0 when it is equal to or less than the dynamic threshold value.
[0151] S2. Perform intra-segment signal clustering on the preprocessed signal to obtain a first cluster signal.
[0152] Furthermore, in step S2, the step of performing intra-segment signal clustering on the preprocessed signal to obtain a first cluster signal includes the following sub-steps:
[0153] S21, obtaining the frequency value of the point power of the first non-zero point in the binary sub-signal, and recording the frequency value in a first cluster storage unit, wherein the first cluster storage unit is a column structure, and the frequency value is stored in the first column of the first cluster storage unit, and at the same time, the frequency value of the point power of the first non-zero point in the binary sub-signal is used as a first reference value;
[0154] S22, obtaining the frequency value of the point power of the next non-zero point in the binarized sub-signal that is not stored in the first cluster storage unit, and performing a difference calculation between the frequency value and the first reference value, wherein:
[0155] If the difference between the frequency value and the first reference value is less than or equal to a first preset frequency clustering threshold, assigning the frequency upper limit in the last non-empty column in the first clustering storage unit with the value of the frequency upper limit of the frequency value;
[0156] If the difference between the frequency value and the first reference value is greater than the first preset frequency clustering threshold, assigning the value of the frequency value to the first reference value, and at the same time, recording the sub-signal corresponding to the frequency value into the first empty column of the first cluster storage unit;
[0157] S23, repeating step S22 until the difference comparison between all the point powers in the binarized sub-signal and the first reference value is completed, and then the first cluster storage unit is output as the first cluster signal.
[0158] S3. Perform inter-segment signal clustering on the first cluster signal to obtain a second cluster signal.
[0159] Furthermore, in step S3, the step of performing inter-segment signal clustering on the first cluster signal to obtain a second cluster signal includes the following sub-steps:
[0160] S31, taking the frequency value and time value of the first column in the first cluster signal as the second reference value, and recording the sub-signal corresponding to the first column in the first cluster signal into a second cluster storage unit, the second cluster storage unit being a column structure, wherein the second column in the first cluster signal is taken as the current column of the inter-segment signal clustering;
[0161] S32, comparing the frequency value and the time value of the sub-signal in the current column with the second reference value, wherein:
[0162] If the difference between the frequency value and the second reference value is less than the second preset clustering threshold, and the difference between the time value and the second reference value is less than the preset time clustering threshold, then continue to determine whether the upper frequency limit in the frequency value is greater than the second reference value, wherein if so, assign the value of the upper frequency limit to the second reference value; if not, keep the second reference value unchanged;
[0163] If the difference between the frequency value and the second reference value is greater than or equal to the second preset clustering threshold, or the difference between the time value and the second reference value is greater than or equal to the preset time clustering threshold, storing the sub-signal corresponding to the current column into the first empty column in the second clustering storage unit;
[0164] After the difference between the frequency value and the time value of the current column and the second reference value is compared, the next column of the current column is used as the new current column;
[0165] S33, repeating step S32 until all the sub-signals in the first cluster signal are compared with the second reference value, and then outputting all the sub-signals in the second cluster storage unit as the second cluster signal.
[0166] S4. Perform inter-cluster signal clustering on the second cluster signal to obtain a third cluster signal.
[0167] Furthermore, in step S4, the step of performing inter-class signal clustering on the second cluster signal to obtain a third cluster signal includes the following sub-steps:
[0168] S41, taking the pulse width of the first column in the second cluster signal as a third reference value, and recording the sub-signal corresponding to the first column in the second cluster signal into a third cluster storage unit, wherein the third cluster storage unit is a column structure, wherein the second column in the second cluster signal is used as a current column for clustering inter-class signals;
[0169] S42, comparing the pulse width of the sub-signal in the current column with the third reference value, wherein:
[0170] If the difference between the pulse width and the third reference value is smaller than a preset pulse width clustering threshold, storing the sub-signal corresponding to the current column into a first empty column of the third clustering storage unit;
[0171] If the difference between the pulse width and the third reference value is greater than or equal to the preset pulse width clustering threshold, assigning the value of the pulse width to the third reference value;
[0172] After the difference between the pulse width of the current column and the third reference value is compared, the next column of the current column is used as the new current column;
[0173] S43, repeating step S42 until all the sub-signals in the second cluster signal are compared with the third reference value, and then outputting all the sub-signals in the third cluster storage unit as the third cluster signal.
[0174] S5. Classify and identify the third cluster signal to obtain a drone detection result.
[0175] Furthermore, in step S5, the step of classifying and distinguishing the third cluster signal to obtain the drone detection result includes the following sub-steps:
[0176] S51, taking out all the sub-signals in the third clustering signal to obtain a one-dimensional array of S columns, subtracting the parameter values of the sub-signals in the first column of the one-dimensional array from the parameter values corresponding to all columns except the first column, and at the same time, subtracting the parameter values of the sub-signals in the second column of the one-dimensional array from the parameter values corresponding to all columns except the first column and the second column to obtain (S-1)*S / 2 feature judgment difference values, and grouping the feature judgment difference values according to a preset feature judgment threshold to obtain N difference value groups;
[0177] S52, the feature judgment difference corresponding to the group containing the largest number of feature judgment difference values in the difference grouping is recorded as the pulse interval time, and the pulse width and the pulse interval time of every two adjacent sub-signals in the one-dimensional array are respectively calculated with the preset drone judgment signal, wherein:
[0178] If the difference between the pulse width and the preset drone judgment signal is less than or equal to the preset feature judgment difference, or the difference between the pulse interval time and the preset drone judgment signal is less than or equal to the preset feature judgment difference, it is determined that there is a drone signal between the corresponding two adjacent sub-signals;
[0179] If the difference between the pulse width or the inter-pulse interval time of the sub-signal and the preset drone judgment signal is greater than the preset feature judgment difference, it is determined that there is no drone signal between the corresponding two adjacent sub-signals.
[0180] The computer device 300 provided in the embodiment of the present invention can implement the steps in the drone detection method based on software radio in the above embodiment, and can achieve the same technical effect. Please refer to the description in the above embodiment and will not be repeated here.
[0181] Exemplarily, a hardware that can implement the software radio-based drone detection method in an embodiment of the present invention is a software radio platform based on AD9361, and the software radio platform includes an FPGA for signal processing and classification, a software radio chip AD9361 for collecting signals, and an STM32 chip for configuring the parameters of the software radio chip AD9361 and storing and reading the drone identification library.
[0182] An embodiment of the present invention further provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the various processes and steps in the software radio-based drone detection method provided in the embodiment of the present invention are implemented, and the same technical effect can be achieved. To avoid repetition, it will not be repeated here.
[0183] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiments can be implemented by instructing the relevant hardware through a computer program, and the program can be stored in a computer-readable storage medium, and when the program is executed, it can include the processes of the embodiments of the above-mentioned methods. The storage medium can be a disk, an optical disk, a read-only memory (ROM) or a random access memory (RAM).
[0184] It should be noted that, in this article, the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, an element defined by the sentence "comprises a ..." does not exclude the existence of other identical elements in the process, method, article or device including the element.
[0185] Through the description of the above implementation methods, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of software plus a necessary general hardware platform, and of course by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, a magnetic disk, or an optical disk), and includes a number of instructions for enabling a terminal (which can be a mobile phone, a computer, a server, an air conditioner, or a network device, etc.) to execute the methods described in each embodiment of the present invention.
[0186] The embodiments of the present invention are described above in conjunction with the accompanying drawings. What is disclosed is only the preferred embodiment of the present invention. However, the present invention is not limited to the above-mentioned specific implementation manner. The above-mentioned specific implementation manner is only illustrative rather than restrictive. Under the enlightenment of the present invention, ordinary technicians in this field can also make many forms and equivalent changes without departing from the scope of protection of the purpose of the present invention and the claims, all of which are within the protection of the present invention.
Claims
1. A drone detection method based on software radio, characterized in that: The method comprises the following steps: S1. Acquire a radio signal and preprocess the radio signal to obtain a preprocessed signal, wherein the parameters of the radio signal include a power value, a frequency value, a time value, a bandwidth and a pulse width, the frequency value includes an upper frequency limit and a lower frequency limit, and the time value includes an arrival time and an end time; Step S1 includes the following sub-steps: S11, dividing the radio signal into sub-signal lengths M Segmenting to obtain multiple sub-signals, wherein the sub-signals at least include a continuous first sub-signal and a second sub-signal, and performing Fourier transform processing on each sub-signal; S12, performing dynamic denoising on the sub-signal after Fourier transformation, wherein the sub-signal has a length M The power of each point in y , calculate the maximum power of each sub-signal in its frequency domain y max , and calculate the power mean value based on the maximum power value y mean ; S13. Judgment y max / y mean Whether the value is less than the preset threshold value, and set the dynamic threshold value; S14, binarizing the power value of the sub-signal according to the dynamic threshold value to obtain a binarized sub-signal; S2, performing intra-segment signal clustering on the preprocessed signal to obtain a first cluster signal; Step S2 includes the following sub-steps: S21, obtaining the frequency value of the point power of the first non-zero point in the binary sub-signal, and recording the frequency value in a first cluster storage unit, wherein the first cluster storage unit is a column structure, and the frequency value is stored in the first column of the first cluster storage unit, and at the same time, the frequency value of the point power of the first non-zero point in the binary sub-signal is used as a first reference value; S22, obtaining the frequency value of the point power of the next non-zero point in the binarized sub-signal that is not stored in the first cluster storage unit, and performing a difference calculation between the frequency value and the first reference value, wherein: If the difference between the frequency value and the first reference value is less than or equal to a first preset frequency clustering threshold, assigning the frequency upper limit in the last non-empty column in the first clustering storage unit with the value of the frequency upper limit of the frequency value; If the difference between the frequency value and the first reference value is greater than the first preset frequency clustering threshold, assigning the value of the frequency value to the first reference value, and at the same time, recording the sub-signal corresponding to the frequency value into the first empty column of the first cluster storage unit; S23, repeating step S22 until the difference between all the point powers in the binarized sub-signal and the first reference value is compared, and then the first cluster storage unit is output as the first cluster signal; S3, performing inter-segment signal clustering on the first cluster signal to obtain a second cluster signal; Step S3 includes the following sub-steps: S31, taking the frequency value and time value of the first column in the first cluster signal as the second reference value, and recording the sub-signal corresponding to the first column in the first cluster signal into a second cluster storage unit, the second cluster storage unit being a column structure, wherein the second column in the first cluster signal is taken as the current column of the inter-segment signal clustering; S32, comparing the frequency value and the time value of the sub-signal in the current column with the second reference value, wherein: If the difference between the frequency value and the second reference value is less than the second preset clustering threshold, and the difference between the time value and the second reference value is less than the preset time clustering threshold, then continue to determine whether the upper frequency limit in the frequency value is greater than the second reference value, wherein if so, assign the value of the upper frequency limit to the second reference value; if not, keep the second reference value unchanged; If the difference between the frequency value and the second reference value is greater than or equal to the second preset clustering threshold, or the difference between the time value and the second reference value is greater than or equal to the preset time clustering threshold, storing the sub-signal corresponding to the current column into the first empty column in the second clustering storage unit; After the difference between the frequency value and the time value of the current column and the second reference value is compared, the next column of the current column is used as the new current column; S33, repeating step S32 until all the sub-signals in the first cluster signal are compared with the second reference value, and then outputting all the sub-signals in the second cluster storage unit as the second cluster signal; S4, performing inter-class signal clustering on the second clustering signal to obtain a third clustering signal; Step S4 includes the following sub-steps: S41, taking the pulse width of the first column in the second cluster signal as a third reference value, and recording the sub-signal corresponding to the first column in the second cluster signal into a third cluster storage unit, wherein the third cluster storage unit is a column structure, wherein the second column in the second cluster signal is used as a current column for clustering inter-class signals; S42, comparing the pulse width of the sub-signal in the current column with the third reference value, wherein: If the difference between the pulse width and the third reference value is smaller than a preset pulse width clustering threshold, storing the sub-signal corresponding to the current column into a first empty column of the third clustering storage unit; If the difference between the pulse width and the third reference value is greater than or equal to the preset pulse width clustering threshold, assigning the value of the pulse width to the third reference value; After the difference between the pulse width of the current column and the third reference value is compared, the next column of the current column is used as the new current column; S43, repeating step S42 until all the sub-signals in the second cluster signal are compared with the third reference value, and outputting all the sub-signals in the third cluster storage unit as the third cluster signal; S5. Classify and identify the third cluster signal to obtain a drone detection result; Step S5 includes the following sub-steps: S51, taking out all the sub-signals in the third cluster signal, obtaining S The parameter value of the sub-signal in the first column of the one-dimensional array is subtracted from the parameter values corresponding to all columns except the first column to form S-1 feature judgment difference values. At the same time, the parameter value of the sub-signal in the second column of the one-dimensional array is subtracted from the parameter values corresponding to all columns except the first column and the second column to form S-2 feature judgment difference values. And so on, the result is (S-1)*S / 2 feature judgment difference values, and grouping the feature judgment difference values according to a preset feature judgment threshold to obtain N Difference groups; S52, the feature judgment difference corresponding to the group containing the largest number of feature judgment difference values in the difference grouping is recorded as the pulse interval time, and the pulse width and the pulse interval time of every two adjacent sub-signals in the one-dimensional array are respectively calculated with the preset drone judgment signal, wherein: If the difference between the pulse width and the preset drone judgment signal is less than or equal to the preset feature judgment difference, or the difference between the pulse interval time and the preset drone judgment signal is less than or equal to the preset feature judgment difference, it is determined that there is a drone signal between the corresponding two adjacent sub-signals; If the difference between the pulse width and the inter-pulse interval time of the sub-signal and the preset drone judgment signal is greater than the preset feature judgment difference, it is determined that there is no drone signal between the corresponding two adjacent sub-signals.
2. The method for detecting unmanned aerial vehicles based on software defined radio as claimed in claim 1, characterized in that: In step S13: like y max / y mean If the value of is less than the preset threshold value, the value of the dynamic threshold value is set to y max +1 ; like y max / y mean is equal to or greater than the preset threshold value, the value of the dynamic threshold value is set to y mean The value obtained by multiplying the preset threshold value.
3. The software radio-based drone detection method according to claim 2, characterized in that: The signal overlap between the consecutive sub-signals is 50% each, wherein for a length of M In the sub-signal, the point power of each point y There is a maximum power value y max , the maximum power y max The following relationship (1) is satisfied: y max =max{y 2 (k)},k∈(0,1,…,M-1) (1); For the point power, the point powers of the first sub-signal and the second sub-signal are defined as follows: y1 and y2 , the length is M The sub-signals within the range are calculated separately y1 / y2 , and the length is M New point power y new , among which, if y1 / y2 If the power is greater than or equal to the preset power threshold, y1 The new point power mark of the start is set clear_flag ,until y1 / y2 is greater than the preset power threshold value, and the new point power satisfies the following relationship (2): (2); The power mean y mean The value is equal to y(k) new Add and divide by y(k) new The number of non-zero values in ; The preset threshold is defined as sel_th , the dynamic threshold is noise_th , then the dynamic threshold value noise_th The following relationship (3) is satisfied: (3); According to the dynamic threshold value, each point power of the sub-signal y Perform binarization processing to obtain the binarized sub-signal, wherein, y When it is greater than the dynamic threshold value, it is set to 1. y When the dynamic threshold value is equal to or less than the dynamic threshold value, it is set to 0.
4. A drone detection system based on software radio, characterized in that: include: The signal preprocessing module is used to acquire a radio signal and preprocess the radio signal to obtain a preprocessed signal. The parameters of the radio signal include a power value, a frequency value, a time value, a bandwidth and a pulse width. The frequency value also includes an upper frequency limit and a lower frequency limit. The time value also includes an arrival time and an end time. The signal preprocessing module is specifically used to perform the following steps: S11, dividing the radio signal into sub-signal lengths M Segmenting to obtain multiple sub-signals, wherein the sub-signals at least include a continuous first sub-signal and a second sub-signal, and performing Fourier transform processing on each sub-signal; S12, performing dynamic denoising on the sub-signal after Fourier transformation, wherein the sub-signal has a length M The power of each point in y , calculate the maximum power of each sub-signal in its frequency domain y max , and calculate the power mean value based on the maximum power value y mean ; S13. Judgment y max / y mean Whether the value is less than the preset threshold value, and set the dynamic threshold value; S14, binarizing the power value of the sub-signal according to the dynamic threshold value to obtain a binarized sub-signal; The intra-segment clustering module is used to perform intra-segment signal clustering on the preprocessed signal to obtain a first cluster signal; the intra-segment clustering module is specifically used to perform the following steps: S21, obtaining the frequency value of the point power of the first non-zero point in the binary sub-signal, and recording the frequency value in a first cluster storage unit, wherein the first cluster storage unit is a column structure, and the frequency value is stored in the first column of the first cluster storage unit, and at the same time, the frequency value of the point power of the first non-zero point in the binary sub-signal is used as a first reference value; S22, obtaining the frequency value of the point power of the next non-zero point in the binarized sub-signal that is not stored in the first cluster storage unit, and performing a difference calculation between the frequency value and the first reference value, wherein: If the difference between the frequency value and the first reference value is less than or equal to a first preset frequency clustering threshold, assigning the frequency upper limit in the last non-empty column in the first clustering storage unit with the value of the frequency upper limit of the frequency value; If the difference between the frequency value and the first reference value is greater than the first preset frequency clustering threshold, assigning the value of the frequency value to the first reference value, and at the same time, recording the sub-signal corresponding to the frequency value into the first empty column of the first cluster storage unit; S23, repeating step S22 until the difference between all the point powers in the binarized sub-signal and the first reference value is compared, and then the first cluster storage unit is output as the first cluster signal; An inter-segment clustering module is used to perform inter-segment signal clustering on the first cluster signal to obtain a second cluster signal; the inter-segment clustering module is specifically used to perform the following steps: S31, taking the frequency value and time value of the first column in the first cluster signal as the second reference value, and recording the sub-signal corresponding to the first column in the first cluster signal into a second cluster storage unit, the second cluster storage unit being a column structure, wherein the second column in the first cluster signal is taken as the current column of the inter-segment signal clustering; S32, comparing the frequency value and the time value of the sub-signal in the current column with the second reference value, wherein: If the difference between the frequency value and the second reference value is less than the second preset clustering threshold, and the difference between the time value and the second reference value is less than the preset time clustering threshold, then continue to determine whether the upper frequency limit in the frequency value is greater than the second reference value, wherein if so, assign the value of the upper frequency limit to the second reference value; if not, keep the second reference value unchanged; If the difference between the frequency value and the second reference value is greater than or equal to the second preset clustering threshold, or the difference between the time value and the second reference value is greater than or equal to the preset time clustering threshold, storing the sub-signal corresponding to the current column into the first empty column in the second clustering storage unit; After the difference between the frequency value and the time value of the current column and the second reference value is compared, the next column of the current column is used as the new current column; S33, repeating step S32 until all the sub-signals in the first cluster signal are compared with the second reference value, and then outputting all the sub-signals in the second cluster storage unit as the second cluster signal; An inter-class clustering module is used to perform inter-class signal clustering on the second clustering signal to obtain a third clustering signal; the inter-class clustering module is specifically used to perform the following steps: S41, taking the pulse width of the first column in the second cluster signal as a third reference value, and recording the sub-signal corresponding to the first column in the second cluster signal into a third cluster storage unit, wherein the third cluster storage unit is a column structure, wherein the second column in the second cluster signal is used as a current column for clustering inter-class signals; S42, comparing the pulse width of the sub-signal in the current column with the third reference value, wherein: If the difference between the pulse width and the third reference value is smaller than a preset pulse width clustering threshold, storing the sub-signal corresponding to the current column into a first empty column of the third clustering storage unit; If the difference between the pulse width and the third reference value is greater than or equal to the preset pulse width clustering threshold, assigning the value of the pulse width to the third reference value; After the difference between the pulse width of the current column and the third reference value is compared, the next column of the current column is used as the new current column; S43, repeating step S42 until all the sub-signals in the second cluster signal are compared with the third reference value, and outputting all the sub-signals in the third cluster storage unit as the third cluster signal; A cluster detection module is used to classify and identify the third cluster signal to obtain a drone detection result; the cluster detection module is specifically used to perform the following steps: S51, taking out all the sub-signals in the third cluster signal, obtaining S The parameter value of the sub-signal in the first column of the one-dimensional array is subtracted from the parameter values corresponding to all columns except the first column to form S-1 feature judgment difference values. At the same time, the parameter value of the sub-signal in the second column of the one-dimensional array is subtracted from the parameter values corresponding to all columns except the first column and the second column to form S-2 feature judgment difference values. And so on, the result is (S-1)*S / 2 feature judgment difference values, and grouping the feature judgment difference values according to a preset feature judgment threshold to obtain N Difference groups; S52, the feature judgment difference corresponding to the group containing the largest number of feature judgment difference values in the difference grouping is recorded as the pulse interval time, and the pulse width and the pulse interval time of every two adjacent sub-signals in the one-dimensional array are respectively calculated with the preset drone judgment signal, wherein: If the difference between the pulse width and the preset drone judgment signal is less than or equal to the preset feature judgment difference, or the difference between the pulse interval time and the preset drone judgment signal is less than or equal to the preset feature judgment difference, it is determined that there is a drone signal between the corresponding two adjacent sub-signals; If the difference between the pulse width and the inter-pulse interval time of the sub-signal and the preset drone judgment signal is greater than the preset feature judgment difference, it is determined that there is no drone signal between the corresponding two adjacent sub-signals.
5. A computer device, characterized in that: include: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the steps of the software radio-based drone detection method as described in any one of claims 1 to 3 are implemented.
6. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by the processor, the steps in the software radio-based drone detection method as described in any one of claims 1 to 3 are implemented.
