A method for detecting a UAV under wireless signal interference

Abstract

The application discloses a kind of unmanned aerial vehicle detection methods under wireless signal interference, comprising: intercepting communication signal, communication signal is decomposed and reconstructed, the sample entropy of reconstructed signal is calculated and whether signal enters is judged, sample entropy is compared with preset judgment threshold, signal bandwidth is calculated and whether it is WiFi signal is judged, and entropy feature is compared with Bluetooth signal and whether unmanned aerial vehicle is detected is judged;The application is by intercepting the communication signal between unmanned aerial vehicle and ground remote control, the signal is reconstructed by three-layer wavelet packet decomposition, and the sample entropy of reconstructed signal is calculated, to judge whether signal enters, and when signal enters, the reconstructed signal is Fourier transformed, the signal bandwidth is calculated, whether it is WiFi signal is judged, if it is not WiFi signal, the modulation feature of reconstructed signal is obtained, to detect whether there is unmanned aerial vehicle, to realize in WiFi and bluetooth device interference to unmanned aerial vehicle is detected, with higher detection rate.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of radio signal processing and unmanned aerial vehicle detection, and particularly relates to a method for detecting unmanned aerial vehicles under wireless signal interference. BACKGROUND

[0002] At present, more and more unmanned aerial vehicles are applied to various scenes. Since unmanned aerial vehicles are cheap and easy to obtain, the actual application of unmanned aerial vehicles includes not only legal applications but also various illegal applications, such as taking pictures secretly and throwing illegal objects. At this time, the use of unmanned aerial vehicles cannot be regulated by legal means only, and in order to avoid the harm caused by illegal use of unmanned aerial vehicles, the use of unmanned aerial vehicles needs to be managed, including detection and defense, and the detection of unmanned aerial vehicles is a basic requirement for the management of unmanned aerial vehicles.

[0003] The existing unmanned aerial vehicle detection methods can be roughly divided into four categories: radar detection, visible light / infrared detection, acoustic frequency detection, and radio detection.

[0004] Radar detection is to detect unmanned aerial vehicles by emitting electromagnetic waves to the target airspace and processing the received echo signals. Radar detection can work all day long, is less affected by weather, has a long detection distance and strong multi-target tracking capability, but radar equipment is expensive, bulky, and lacks concealment, which makes it difficult to deploy in urban environments. In addition, radar detection will cause electromagnetic pollution to the environment and is easily disturbed by ground clutter. Moreover, radar detection is mainly aimed at high-altitude, high-speed, and large-volume target aircraft, and unmanned aerial vehicles belong to "low, small, and slow" targets, which reduces the probability of being detected.

[0005] Visible light / infrared light detection is to detect unmanned aerial vehicles by using the infrared radiation difference between the background and the target through the combination of high-definition visible light cameras or infrared thermal imagers. This method has the advantages of low cost, mature industry chain, strong visualization, and high precision, but it is greatly affected by environmental and meteorological conditions. When the target is small, it is easy to be blocked and difficult to detect, and it is also difficult to distinguish unmanned aerial vehicles from similar-sized objects, so the error rate is high.

[0006] Acoustic frequency detection is to detect the sound generated by the internal components of the unmanned aerial vehicle when it moves through acoustic sensors. This method has the advantages of low cost, good safety, and high precision, but the detection distance is too short, and the noise level of the unmanned aerial vehicle is very low, making it difficult to detect in complex acoustic environments such as cities and airports.

[0007] Radio detection is to monitor the remote control signals or image transmission signals of the unmanned aerial vehicle when it communicates with the ground, and to extract the signal features to determine whether it is a unmanned aerial vehicle communication signal. The outstanding advantage of this method is that it is convenient to use, the equipment is low-cost, and it is not affected by strong light, clouds, obstructions, and the size of the unmanned aerial vehicle. It is an effective supplement to anti-unmanned aerial vehicle systems.

[0008] The existing radio detection methods include:

[0009] DRN-UAV algorithm: It is an algorithm for identifying and detecting the time-frequency spectrum of the remote control signal of the unmanned aerial vehicle by using a residual neural network. A large number of measured different model remote control signal test spectrum are used as a data set to train and test the residual neural network, and finally the trained network is used to identify whether the current remote control signal exists and its model in real time;

[0010] Unmanned aerial vehicle detection and identification system based on flight control signal spectrum characteristics: The system takes FPGA as the control and operation core, adopts AD radio frequency transceiver signal detection and identification algorithm processing, and completes the detection and identification of the unmanned aerial vehicle;

[0011] Unmanned aerial vehicle identification based on RF-DNA: The method extracts the fingerprint features contained in the transient part of the unmanned aerial vehicle signal, performs feature dimension reduction by principal component analysis, and finally classifies and identifies the signal by using a multi-class support vector algorithm.

[0012] However, the existing radio detection methods rarely consider the interference of WiFi and Bluetooth signals in the communication frequency band of the unmanned aerial vehicle, and the current papers and patents related to unmanned aerial vehicle detection technology are also based on the WiFi signal characteristics between the unmanned aerial vehicle and the remote controller, which leads to the problem that the existing unmanned aerial vehicle method is not accurate enough for detecting the unmanned aerial vehicle. Therefore, the present application proposes a method for detecting unmanned aerial vehicles under wireless signal interference to solve the problems existing in the prior art. SUMMARY

[0013] In view of the above problems, the purpose of the present application is to propose a method for detecting unmanned aerial vehicles under wireless signal interference, which solves the problem that the existing radio detection methods rarely consider the interference of WiFi and Bluetooth signals in the communication frequency band of the unmanned aerial vehicle, leading to the problem that the existing unmanned aerial vehicle method is not accurate enough for detecting the unmanned aerial vehicle.

[0014] In order to achieve the purpose of the present application, the present application realizes the following technical scheme: a method for detecting unmanned aerial vehicles under wireless signal interference, comprising the following steps:

[0015] Step 1: Use a general software radio device to intercept the communication signals between the unmanned aerial vehicle and the unmanned aerial vehicle remote controller on the ground;

[0016] Step 2: First, the communication signals intercepted by the general software radio device in step 1 are subjected to three-layer wavelet packet decomposition, and then the signals obtained by three-layer wavelet packet decomposition are reconstructed to obtain the reconstructed signals after wavelet packet decomposition;

[0017] Step 3: Calculate the sample entropy of the reconstructed signal based on the reconstructed signal obtained in step 2, and determine whether a signal enters according to the sample entropy calculation result;

[0018] Step four: based on the sample entropy calculated in step three, compare the sample entropy with the preset judgment threshold, if the comparison result indicates that the sample entropy is greater than the preset judgment threshold, it is determined that the wireless signal is detected, and the influence of noise interference is eliminated;

[0019] Step five: first, based on the reconstructed signal obtained in step two, Fourier transform is performed to obtain the highest frequency component and the lowest frequency component of the signal, then the highest frequency component and the lowest frequency component of the signal are subtracted and the signal bandwidth is calculated, and then it is judged whether it is a WiFi signal according to the signal bandwidth calculation result;

[0020] Step six: first, based on the reconstructed signal obtained in step two, the WEE and PSE unmanned aerial vehicle signal entropy feature extraction is performed, the unmanned aerial vehicle signal entropy feature is extracted, then the entropy feature is compared with the Bluetooth signal, and whether the unmanned aerial vehicle is detected is judged according to the comparison result.

[0021] Further improvement lies in that in step two, the specific steps of reconstructing the decomposed signal are: first, calculating the signal energy value of the third layer of 8 nodes after wavelet packet decomposition, then selecting the first three sub-nodes with the largest energy, and reconstructing the signal corresponding to the sub-nodes to obtain the reconstructed signal.

[0022] Further improvement lies in that in step three, the specific calculation steps of sample entropy are: let the reconstructed signal be x, the sequence length be N, and the m-dimensional vector be:

[0023] X(i)=[x(i),x(i+1),...,x(i+m-1)],i∈[1,N-m+1]

[0024] The distance between X(i) and X(j) is defined as:

[0025]

[0026] The similarity tolerance is defined as:

[0027] r=k·SD

[0028] Wherein, k is the tolerance coefficient, SD is the standard deviation of the time series, and the number of all d[X(i),X(j)]<r is recorded as B i , and B i is compared with the total number of vectors N-m+1, and is recorded as:

[0029]

[0030] Increase the dimension to m+1, construct a group of m+1-dimensional vectors, repeat the steps, and obtain:

[0031]

[0032] The expression of the sequence sample entropy is:

[0033]

[0034] Wherein, SampEn (m, r) is sample entropy.

[0035] Further improvement lies in that in the step five, the signal bandwidth calculation formula is:

[0036] B=f max -f min

[0037] In the formula, f max is the highest frequency component, and f min is the lowest frequency component.

[0038] Further improvement lies in that in the step five, in the signal bandwidth calculation result judgment process, if the calculated signal bandwidth is greater than 20 MHz, it is indicated that the signal is a WiFi signal, and the step one is returned to re-perform signal acquisition.

[0039] Further improvement lies in that in the step five, in the signal bandwidth calculation result judgment process, if the calculated signal bandwidth is less than 20 MHz, it is indicated that the signal is not a WiFi signal, and the step six is continuously executed.

[0040] Further improvement lies in that in the step six, in the entropy feature comparison process with the Bluetooth signal, if the extracted unmanned aerial vehicle signal entropy matches the Bluetooth signal feature, it is indicated that there is no unmanned aerial vehicle, and if the extracted unmanned aerial vehicle signal entropy does not match the Bluetooth signal feature, it is indicated that the unmanned aerial vehicle is successfully detected.

[0041] Further improvement lies in that if the reconstructed signal is neither a WiFi signal nor a Bluetooth signal, the signal is an unmanned aerial vehicle signal, indicating that the unmanned aerial vehicle is detected.

[0042] The beneficial effects of the present application are that: the present application intercepts the communication signal between the unmanned aerial vehicle and the ground remote control, and performs three-layer wavelet packet decomposition on the signal, calculates the energy value of the signal after wavelet packet decomposition, reconstructs the signal corresponding to the first three energy larger subbands and calculates the sample entropy of the reconstructed signal, to judge whether the signal enters or not, if the signal enters, the reconstructed signal is subjected to Fourier transform, the signal bandwidth is calculated, whether it is a WiFi signal is judged, if it is not a WiFi signal, the modulation feature of the reconstructed signal is obtained, if the feature matches the Bluetooth signal feature, it is indicated that the unmanned aerial vehicle is not detected, otherwise, it is indicated that the unmanned aerial vehicle is successfully detected, so that the unmanned aerial vehicle is detected under the interference of WiFi and Bluetooth devices, has a high detection rate, improves the unmanned aerial vehicle detection accuracy, and is convenient for engineering implementation. BRIEF DESCRIPTION OF DRAWINGS

[0043] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0044] Figure 1 This is a schematic diagram of the drone detection method of the present invention;

[0045] Figure 2 This is a schematic diagram illustrating the application scenario of the drone detection method of the present invention. Detailed Implementation

[0046] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0047] See Figure 1 , Figure 2 This embodiment provides a method for detecting drones under wireless signal interference, including the following steps:

[0048] Step 1: First, prepare the corresponding general software radio equipment in advance according to the actual signal acquisition needs, and then use the general software radio equipment to intercept the communication signal between the drone and the drone remote controller on the ground.

[0049] Step 2: First, perform three-layer wavelet packet decomposition on the communication signal intercepted by the general software radio device in Step 1. Then, calculate the signal energy values ​​of the eight nodes in the third layer after wavelet packet decomposition. Next, select the three child nodes with the largest energy and reconstruct the signal corresponding to the child node to obtain the reconstructed signal after wavelet packet decomposition.

[0050] Step 3: Calculate the sample entropy of the reconstructed signal based on the signal obtained in Step 2, and determine whether a signal has entered based on the sample entropy calculation result. The specific steps for calculating the sample entropy are as follows: Let the reconstructed signal be x, the sequence length be N, and its constituent m-dimensional vector be:

[0051] X(i)=[x(i),x(i+1),...,x(i+m-1)],i∈[1,N-m+1]

[0052] Define the distance between X(i) and X(j) as:

[0053]

[0054] Define the similarity tolerance, that is:

[0055] r=kSD

[0056] Where k is the tolerance coefficient, SD is the standard deviation of the time series, and the number of all d[X(i), X(j)]<r is recorded as B i , and B i is divided by the total number of vectors N-m+1, denoted as:

[0057]

[0058] Increase the dimension to m+1, construct a set of m+1-dimensional vectors, repeat the steps, and get:

[0059]

[0060] Then the expression of the sample entropy of the sequence is:

[0061]

[0062] Where SampEn(m, r) is the sample entropy;

[0063] Step four: based on the sample entropy calculated in step three, preset the judgment threshold δ, and compare the sample entropy with the preset threshold, if the comparison result indicates that the sample entropy is greater than the preset judgment threshold, it is determined that the wireless signal is detected, and the influence of noise interference is eliminated;

[0064] Step five: first, based on the reconstructed signal obtained in step two, Fourier transform is carried out:

[0065]

[0066] Get the highest frequency component f max and the lowest frequency component f min of the signal, and then subtract the highest frequency component from the lowest frequency component:

[0067] B=f max -f min

[0068] Calculate the signal bandwidth B, and then determine whether it is a WiFi signal according to the signal bandwidth calculation result. In the signal bandwidth calculation result determination process, if the calculated signal bandwidth is greater than 20MHz, it indicates that it is a WiFi signal, and returns to step one to reacquire the signal, if the calculated signal bandwidth is less than 20MHz, it indicates that it is not a WiFi signal, and continues to execute step six;

[0069] Step six: first based on the reconstructed signal obtained in step two, the WEE and PSE unmanned aerial vehicle signal entropy feature extraction is carried out, the unmanned aerial vehicle signal entropy feature is extracted, then the entropy feature is compared with the Bluetooth signal, and whether the unmanned aerial vehicle is detected is judged according to the comparison result;

[0070] The energy corresponding to the kth layer of the reconstructed signal is:

[0071]

[0072] Wherein, A k，i is the low frequency component of the signal at the i th sampling point, D k，i is the high frequency component of the signal at the i th sampling point, and the total energy of the wireless signal is:

[0073]

[0074] The probability of the ratio of the energy of each frequency band to the total energy is:

[0075]

[0076] The corresponding wavelet energy entropy can be expressed as:

[0077]

[0078] For the discrete sequence U of the wireless signal r u (t)={u1,u2,...,u N}, the power spectrum is expressed as:

[0079]

[0080] Wherein, U(w) is the discrete Fourier transform of U, and the probability is calculated by S(w):

[0081]

[0082] In the process of comparing the entropy feature with the Bluetooth signal, if the extracted unmanned aerial vehicle signal entropy matches the Bluetooth signal feature, it indicates that there is no unmanned aerial vehicle, and if the extracted unmanned aerial vehicle signal entropy does not match the Bluetooth signal feature, it indicates that the unmanned aerial vehicle is successfully detected.

[0083] If the noise interference is excluded, and after judgment, the reconstructed signal is neither WiFi signal nor Bluetooth signal, the signal is unmanned aerial vehicle signal, indicating that the unmanned aerial vehicle is detected.

[0084] The above only describes the preferred embodiments of the present application, and does not limit the present application, any modification, equivalent replacement, improvement, etc. within the spirit and principles of the present application, should be included in the protection scope of the present application.

Claims

1. A method for detecting a UAV in a wireless signal interference, characterized in that, The method comprises the following steps: Step 1: intercepting the communication signal between the unmanned aerial vehicle and the unmanned aerial vehicle remote controller on the ground by using a general software radio; Step 2: performing three-layer wavelet packet decomposition on the communication signal intercepted by the general software radio in step 1, and then reconstructing the signal obtained by the three-layer wavelet packet decomposition to obtain a reconstructed signal after wavelet packet decomposition; Step 3: calculating the sample entropy of the reconstructed signal obtained in step 2, and judging whether a signal enters according to the sample entropy calculation result; Step 4: comparing the sample entropy with a preset judgment threshold based on the sample entropy calculated in step 3, and if the comparison result indicates that the sample entropy is greater than the preset judgment threshold, it is determined that a wireless signal is detected, and the influence of noise interference is eliminated; Step 5: performing Fourier transform on the reconstructed signal obtained in step 2 to obtain the highest frequency component and the lowest frequency component of the signal, then subtracting the highest frequency component from the lowest frequency component and calculating the signal bandwidth, and then judging whether it is a WiFi signal according to the signal bandwidth calculation result; Step 6: first, based on the reconstructed signal obtained in step 2, the WEE and PSE unmanned aerial vehicle signal entropy feature extraction is performed to extract the unmanned aerial vehicle signal entropy feature, then the entropy feature is compared with the Bluetooth signal, and whether the unmanned aerial vehicle is detected is judged according to the comparison result. 2.The method of claim 1, wherein: In step 2, the specific steps of reconstructing the decomposed signal are: first, calculating the signal energy value of the third layer of 8 nodes after wavelet packet decomposition, then selecting the first three nodes with the largest energy, and reconstructing the signal corresponding to the nodes to obtain the reconstructed signal. 3.The method of claim 1, wherein: In step 3, the specific calculation steps of the sample entropy are: let the reconstructed signal be x, the sequence length be N, and the m-dimensional vector be: X(i)=[x(i),x(i+1),...,x(i+m-1)],i∈[1,N-m+1] The distance between X(i) and X(j) is defined as: The similarity tolerance is defined as: r=k·SD where k is the tolerance factor, SD is the standard deviation of the time series, and B is the number of all d[X(i), X(j)] < r i and B i is divided by the total number of vectors N - m + 1, and is denoted as: Increase the dimension to m+1, construct a group of m+1-dimensional vectors, repeat the steps, and obtain: The expression of the sequence sample entropy is: Where SampEn(m,r) is the sample entropy. 4.The method of claim 1, wherein: In step 5, the signal bandwidth calculation formula is: B = f max -f min where f max is the highest frequency component, f min is the lowest frequency component.

5. The method of claim 1, wherein: In the signal bandwidth calculation result judgment process in step 5, if the calculated signal bandwidth is greater than 20MHz, it indicates that it is a WiFi signal, and the signal collection is restarted in step 1.

6. The method of claim 1, wherein: In the signal bandwidth calculation result judgment process in step 5, if the calculated signal bandwidth is less than 20MHz, it indicates that it is not a WiFi signal, and step 6 is continued.

7. The method of claim 1, wherein: In step 6, in the process of comparing the entropy feature with the Bluetooth signal, if the extracted unmanned aerial vehicle signal entropy matches the Bluetooth signal feature, it indicates that there is no unmanned aerial vehicle, and if the extracted unmanned aerial vehicle signal entropy does not match the Bluetooth signal feature, it indicates that the unmanned aerial vehicle is successfully detected. 8.The method of claim 1, wherein: If the reconstructed signal is neither a WiFi signal nor a Bluetooth signal, the signal is an unmanned aerial vehicle signal, indicating that the unmanned aerial vehicle is detected.

Images