A fast detection method of UAV frequency hopping signal based on FPGA

Abstract

The present invention discloses a method for quickly detecting frequency hopping signals of unmanned aerial vehicles based on FPGA. The method realizes the functions of short-time Fourier transform and channelization detection by using a sliding DFT algorithm, simplifies the algorithm complexity and saves processing resources. The method saves DSP resources of FPGA by using a fast amplitude estimation algorithm, reduces processing delay and increases response speed. The method eliminates WIFI signal interference in the same frequency band by using real-time calculation of dynamic thresholds. The method eliminates burst signal interference in the same frequency band by using a frequency hopping signal duration judgment mechanism, ensures the accuracy of interference guidance, thereby saving system power consumption and solving the problem of electromagnetic environment pollution caused by invalid interference signals.

Description

Technical Field

[0001] The present invention relates to the field of communication countermeasure technology, and in particular to a method for quickly detecting frequency hopping signals of unmanned aerial vehicles based on FPGA. Background Art

[0002] In the field of civil drone control, most products on the market use high-power / multi-carrier / comb spectrum blind jamming signals. This jamming method is not only ineffective in achieving the expected effect, but also not conducive to the cost control of civil-grade control equipment. It also causes spectrum pollution in specific airspaces. In order to solve the above problems and achieve accurate jamming of drone frequency hopping signals, it is necessary to quickly detect and learn its frequency hopping signals and obtain its frequency hopping patterns in real time. Summary of the invention

[0003] In order to solve the problem of accurate interference of the above-mentioned UAV frequency hopping signal, the present invention proposes a method for rapid detection of UAV frequency hopping signals based on FPGA.

[0004] A method for quickly detecting frequency hopping signals of unmanned aerial vehicles based on FPGA includes the following steps:

[0005] Step S1: Collect broadband input signals, perform channelized detection processing, and output complex spectrum data of multiple channels;

[0006] Step S2: Perform rapid amplitude calculation on the output complex spectrum data to obtain a real-time spectrum;

[0007] Step S3: dynamically calculate the decision threshold to eliminate fixed-frequency interference in the same frequency band;

[0008] Step S4: Determine the duration of the UAV frequency hopping signal and eliminate the burst interference in the same frequency band.

[0009] Furthermore, the channelized detection processing in step S1 is based on sliding DFT, and divides the channel detection into a comb filter circuit of a common part and an integration circuit of each channel; and also includes dividing the detection channel according to the resource planning of FPGA.

[0010] Furthermore, the time domain real sequence to be detected based on the sliding DFT channelization detection input is x(n), and the spectrum output of the mth channel is the complex sequence X m (n), the number of divided channels is N, the value range of m is 0~N / 2-1, the frequency resolution is fs / N, and the frequency of the mth channel detection is m*fs / N.

[0011] Further, the step S2 of rapid amplitude calculation includes the following sub-steps:

[0012] Step S21: firstly, perform an absolute value operation on the spectrum signal;

[0013] Step S22: Find the maximum and minimum values ​​of the spectrum absolute value

[0014] Step S23: multiply the large value by the coefficient 1, multiply the small value by the coefficient 2, and add them together to output.

[0015] Furthermore, the coefficient 1 is preferably 0.99, and the coefficient 2 is preferably 0.197.

[0016] Further, the step S3 of dynamically calculating the decision threshold includes:

[0017] Step S31: Select a time twice the length of the drone signal as a detection window, extract and accumulate the amplitude signal output in step S2 within the detection window, and calculate the average value;

[0018] Step S32: Calculate the difference between the extracted signal and the average value. If the difference is positive, it is judged as 1, and if the difference is negative, it is judged as 0.

[0019] Furthermore, the step S3 of dynamically calculating the decision threshold also includes: in order to reduce the storage resources of the FPGA, downsampling is performed on the basis of the original sampling rate, that is, only some points are taken within the detection window for calculation.

[0020] Further, the step S4 determines the duration of the drone frequency hopping signal, including: performing a duration determination on the difference determination result outputted from step S3; if the duration determined as 1 exceeds 2 / 3 of the typical drone frequency hopping residence time, it is determined to be a drone frequency hopping signal, otherwise it is determined to be an interference signal.

[0021] Beneficial effects of the present invention: The present invention proposes a method for rapid detection of frequency-hopping signals of unmanned aerial vehicles based on FPGA, and realizes the functions of short-time Fourier transform and channelized detection by using a sliding DFT algorithm, thereby simplifying the algorithm complexity and saving processing resources; utilizing a fast amplitude estimation algorithm, saving DSP resources of FPGA, while reducing processing delay and increasing response speed; utilizing real-time calculation of dynamic thresholds, eliminating interference from WIFI signals in the same frequency band; utilizing a frequency-hopping signal duration judgment mechanism, eliminating interference from burst signals in the same frequency band, and ensuring the accuracy of interference guidance, thereby saving system power consumption and solving the problem of electromagnetic environment pollution caused by invalid interference signals. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 It is a schematic diagram of the channelized detection processing based on sliding DFT in the present invention;

[0023] Figure 2 is a diagram of frequency hopping channel detection results according to an embodiment of the present invention;

[0024] Figure 3This is a diagram of the detection result of frequency hopping channel plus collision fixed frequency interference in an embodiment of the present invention;

[0025] Figure 4 This is a diagram of the detection result of a non-collision fixed-frequency interference channel according to an embodiment of the present invention;

[0026] Figure 5 2 is a broadband noise diagram of an embodiment of the present invention. DETAILED DESCRIPTION

[0027] In order to have a clearer understanding of the technical features, purposes and effects of the present invention, specific embodiments of the present invention are now described with reference to the accompanying drawings.

[0028] The present invention proposes a method for quickly detecting frequency hopping signals of unmanned aerial vehicles based on FPGA, comprising the following steps:

[0029] Step S1: Collect broadband input signals, perform channelized detection processing, and output complex spectrum data of multiple channels;

[0030] Step S2: Perform rapid amplitude calculation on the output complex spectrum data to obtain a real-time spectrum;

[0031] Step S3: dynamically calculate the decision threshold to eliminate fixed-frequency interference in the same frequency band;

[0032] Step S4: Determine the duration of the UAV frequency hopping signal and eliminate the burst interference in the same frequency band.

[0033] In step S1, the detection channel is divided according to the resource planning of FPGA. The more the detection signal is divided, the stronger the anti-interference performance is, and the more resources are consumed, and vice versa. The channel detection is divided into a comb filter circuit of the common part and an integration circuit of each channel. The specific circuit form is as follows: Figure 1 As shown; use MATLAB to calculate e in advance j2πm / N and store the value.

[0034] In step S2, the complex IQ signal output in step S1 is quickly calculated for amplitude; first, the absolute values ​​of the I and Q signals are taken, and then the large and small values ​​of the I and Q absolute values ​​are found, and finally the large value is multiplied by the coefficient 1, and the small value is multiplied by the coefficient 2 and then added and output; the best example of coefficient 1 is 0.99, and the best example of coefficient 2 is 0.197.

[0035] In step S3, a time period twice the length of the drone signal is selected as the detection window, and the amplitude signal output from step S200 is extracted and accumulated within the detection window, and the average value is calculated; the difference between the extracted signal and the average value is calculated, and a positive difference is judged as 1, and a negative difference is judged as 0.

[0036] In step S4, the duration of the result output in step S3 is determined; if the duration of 1 exceeds 2 / 3 of the typical UAV frequency hopping dwell time, it is determined to be a UAV frequency hopping signal, otherwise it is determined to be an interference signal.

[0037] In this embodiment, the principle diagram of the channelization detection method based on SDFT (sliding DFT) is as follows: Figure 1 As shown, x(n) is the input real sequence to be detected in the time domain, X m (n) is the spectrum output of the mth channel, which is a complex sequence. N is the number of divided channels, and the value of m is 0 to N / 2-1, that is, only N / 2 channels can be detected. The frequency resolution is fs / N. Therefore, the frequency detected by the mth channel is m*fs / N. According to the value of N, e j2πm / N The value of can be calculated in advance by MATLAB and saved as a file, and then read directly when used. This algorithm mainly consumes LUT resources and complex multiplier (DSP48E) resources, and is proportional to the number of channels N; the flexibility lies in that the value of m can be randomly selected for instantiation according to the frequency range of interest to save resources. The comb filter part in the front is a common part and only needs to be instantiated once; there is a delay of N clock cycles for the detection of each signal, and the subsequent pipeline operation can output spectrum information in real time.

[0038] In this embodiment, in order to obtain the amplitude (power value) of the output frequency of the mth channel, it is necessary to calculate X in real time. m (n) is the amplitude value. The conventional approach is to calculate To save multiplier resources, an approximate fitting algorithm can be used, using only shift operations and addition operations: A≈α*max(abs(I), abs(Q)+β*min(abs(I), abs(Q)), where α is 0.990 and β is 0.197, and the approximate effect is the best.

[0039] In this embodiment, the WIFI signal and the drone frequency hopping signal have overlapping spectrum usage, which may interfere with the detection of the frequency hopping signal.

[0040] WIFI signals are modulated using OFDM. Although the number of subcarriers and specific frequencies may change, the frequency will not jump within a certain period of time. Therefore, WIFI signals can be regarded as fixed-frequency interference. The difference between fixed-frequency signals and frequency-hopping signals is that within a certain period of time, the instantaneous power of fixed-frequency signals is equal to the average power, while the instantaneous power of frequency-hopping signals is greater than or less than the average power. Therefore, fixed-frequency interference can be removed by setting a dynamic detection threshold.

[0041] After channelization through SDFT, the time-frequency diagram of the signal can be obtained. Since the collected signal may contain fixed-frequency signals, frequency-hopping signals, short-time burst signals, and noise signals in the band, in order to obtain the best decision threshold, the present invention uses the idea of ​​power cancellation to make a quick decision based on the characteristics of the frequency-hopping signal.

[0042] Set a detection window of a certain length, accumulate power within the detection window, and calculate its average value The principle for determining the length of the detection window is that the frequency hopping signal exists discontinuously within the detection window, while the fixed-frequency interference and noise exist continuously. Therefore, the detection window must be at least longer than the duration of a frequency hopping signal. Here, the detection window takes the duration of two frequency hopping signals, that is, it includes the two periods of time when the frequency hopping signal exists and when it does not exist.

[0043] In order to reduce the storage resources of FPGA, it is necessary to perform downsampling processing based on the original sampling rate, that is, only some points in the detection window are taken for calculation.

[0044] Subtract the mean from the original data and determine the sign of the new data. If the sign is positive, it is judged as 1, and if the sign is negative, it is judged as 0.

[0045] After the above steps, the binary data of the frequency hopping signal can be obtained. However, due to the sudden change of the signal phase or the strong burst signal interference at a certain moment, discontinuous points may appear on the binary graph. In order to further accurately guide the interference signal, the duration of the frequency hopping signal is judged, that is, the frequency hopping signal duration threshold is set. Only continuous signals exceeding the threshold are recorded, and values ​​that do not exceed the threshold are eliminated. The duration threshold can be set to 2 / 3 of the typical frequency hopping period of the drone.

[0046] In the first embodiment:

[0047] (1) Set the ADC sampling rate to 240 MHz, the number of instantiated channels to 240, and the fast amplitude detection parameters α = 0.990, β = 0.197.

[0048] (2) The dwell time of the typical DJI Phantom 3 UAV frequency hopping signal is 3.262ms. 128 points are taken in the time period when the frequency hopping signal exists and does not exist in the detection window, for a total of 256 points, and the downsampling rate is 6000 times.

[0049] (3) Taking 0.6 times the duration of the frequency hopping signal as the effective frequency hopping signal detection threshold, the number of points is 0.6*40KHz*3.262ms=78, that is, 78 consecutive high signals are used as the basis for judging that it is a frequency hopping signal, and less than 78 consecutive high signals are judged as interference.

[0050] (4) Construct a mixed test signal (with a signal-to-noise ratio of -5dB) containing noise interference, non-collision fixed-frequency interference, collision fixed-frequency interference and the frequency-hopping signal to be detected for testing.

[0051] The frequency hopping channel detection results are as follows: Figure 2 As shown, the detection results of non-collision fixed-frequency interference channels are as follows Figure 3 As shown, the detection results of non-collision fixed-frequency interference channels are as follows Figure 4 As shown, broadband noise is Figure 5 shown.

[0052] The present invention proposes a method for rapid detection of frequency hopping signals of unmanned aerial vehicles based on FPGA, which realizes the functions of short-time Fourier transform and channelized detection simultaneously by using a sliding DFT algorithm, simplifies the algorithm complexity and saves processing resources; utilizes a fast amplitude estimation algorithm to save DSP resources of FPGA, reduces processing delay and increases response speed; utilizes real-time calculation of dynamic threshold to eliminate WIFI signal interference in the same frequency band; utilizes a frequency hopping signal duration judgment mechanism to eliminate burst signal interference in the same frequency band, ensures the accuracy of interference guidance, thereby saving system power consumption and solving the problem of electromagnetic environment pollution caused by invalid interference signals.

[0053] The present invention has shown and described the basic principles and main features of the present invention and the advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The above embodiments and descriptions are only for explaining the principles of the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, which fall within the scope of the present invention. The scope of protection of the present invention is defined by the attached claims and their equivalents.

Claims

1. A method for rapid detection of UAV frequency hopping signals based on FPGA, characterized in that: The following steps are involved: Step S1: Collect broadband input signals, perform channelized detection processing, and output complex spectrum data of multiple channels; Step S2: Perform fast amplitude calculation on the output complex spectrum data to obtain a real-time spectrum; the fast amplitude calculation includes the following sub-steps: Step S21: firstly, perform an absolute value operation on the spectrum signal; Step S22: Find the maximum and minimum values ​​of the absolute value of the spectrum; Step S23: multiply the large value by coefficient one, multiply the small value by coefficient two, add them together and output; Step S3: dynamically calculating the decision threshold to eliminate fixed-frequency interference in the same frequency band. The dynamic calculation of the decision threshold includes: Step S31: Select a time twice the length of the drone signal as a detection window, extract and accumulate the amplitude signal output in step S2 within the detection window, and calculate the average value; Step S32: Calculate the difference between the extracted signal and the average value, if the difference is positive, it is judged as 1, and if the difference is negative, it is judged as 0; Step S4: Determine the duration of the UAV frequency hopping signal and eliminate the burst interference in the same frequency band.

2. The method for rapid detection of UAV frequency hopping signals based on FPGA according to claim 1 is characterized in that: The channelized detection processing in step S1 is based on sliding DFT, and divides the channel detection into a comb filter circuit of a common part and an integration circuit of each channel; and also includes dividing the detection channel according to the resource planning of FPGA.

3. The method for rapid detection of UAV frequency hopping signals based on FPGA according to claim 2 is characterized in that: The real time sequence to be detected based on the sliding DFT channelized detection input is x(n), and the spectrum output of the mth channel is the complex sequence x m (n), the number of divided channels is N, the value range of m is 0~N / 2-1, the frequency resolution is fs / N, and the frequency of the mth channel detection is m*fs / N.

4. The method for rapid detection of UAV frequency hopping signals based on FPGA according to claim 1 is characterized in that: The value of the coefficient one is 0.99, and the value of the coefficient two is 0.

197.

5. The method for rapid detection of UAV frequency hopping signals based on FPGA according to claim 1 is characterized in that: The step S3 of dynamically calculating the decision threshold further includes: in order to reduce the storage resources of the FPGA, downsampling is performed on the basis of the original sampling rate, that is, only some points in the detection window are taken for calculation.

6. The method for rapid detection of UAV frequency hopping signals based on FPGA according to claim 1, characterized in that: The step S4 determines the duration of the drone frequency hopping signal, including: performing a duration determination on the difference determination result outputted from step S3; if the duration determined as 1 exceeds 2 / 3 of the typical drone frequency hopping dwell time, it is determined to be a drone frequency hopping signal, otherwise it is determined to be an interference signal.

Images