CN116087887A

CN116087887A - Radar pulse signal detection method and device based on spectrum sparse sensing

Info

Publication number: CN116087887A
Application number: CN202211665768.1A
Authority: CN
Inventors: 李力
Original assignee: Guoke Dianlei Beijing Electronic Equipment Technology Co ltd
Current assignee: Guoke Dianlei Beijing Electronic Equipment Technology Co ltd
Priority date: 2022-12-23
Filing date: 2022-12-23
Publication date: 2023-05-09

Abstract

The invention relates to a radar pulse signal detection method and device based on spectrum sparse sensing. The radar pulse signal detection device based on the spectrum sparse sensing comprises a spectrum sparse sensing module, a dynamic adjustment self-adaptive filtering module and a double-threshold pulse detection module, wherein the spectrum sparse sensing module is used for carrying out single-bit sampling on an original signal and obtaining signal frequency and signal bandwidth, and the dynamic adjustment self-adaptive filtering module is used for filtering the signal according to the signal frequency and the signal bandwidth. The symbol bits of the AD data are utilized for spectrum sensing, and the spectrum distribution condition of the input is monitored and tracked in real time, so that compared with the traditional short-time Fourier transform, the signal processing resource can be greatly reduced, and the processing bandwidth of the signal processing can be improved. During filtering, according to the central frequency and the bandwidth of the pulse signal obtained by the frequency spectrum monitoring module, a dynamic adjustment self-adaptive filtering module is used for adjusting the central frequency point of the filter, so that tracking filtering of the radar pulse signal is realized.

Description

Radar pulse signal detection method and device based on spectrum sparse sensing

Technical Field

The invention relates to the technical field of radar pulse signal reconnaissance and interception in electronic reconnaissance, in particular to a radar pulse signal detection method and device based on spectrum sparse sensing.

Background

A large number of radar radiation source signals exist in the range of 30 MHz-18 GHz, and the frequency point, bandwidth and signal pattern of each signal are different, so that a very complex electromagnetic environment in the current society is formed. Electromagnetic equipment used by various armies on a battlefield is continuously improved in electromagnetic signal interception resistance, and higher requirements are put on the system performance of a reconnaissance receiver, such as wider monitoring frequency bands, larger dynamic range, higher sensitivity and the like. The growing commercial spectrum demands, sharing and congestion of spectrum resources, have made the electromagnetic spectrum environment faced by radar reconnaissance receivers increasingly complex. The frequency domain wide open channelized receiver has the characteristics of high interception probability, high sensitivity, high frequency resolution and adaptability to simultaneous arrival signals, and the performance characteristics of the frequency domain wide open channelized receiver can effectively cope with modern high-density signal environments. In recent years, the sampling frequency of an analog-to-digital converter (ADC) is rapidly increased, and the highest sampling rate is more than 10 GHz. High-speed ADCs, while greatly widening the instantaneous bandwidth of digital receivers, also present a greater challenge to digital storage and processing capabilities. The channelizing processing of the digital receiver is mainly realized by a programmable logic array (FPGA) chip and a special program. At a sampling rate of 10GHz, the traditional digital channelized processing is difficult to realize at the capacity and the speed of the current FPGA. Therefore, how to further reduce the resource occupation and processing power consumption of the FPGA and to quickly and efficiently detect the radar signal with low signal-to-noise ratio and complex system is a challenging problem.

The Chinese patent with the grant publication number of CN113447893B discloses a radar pulse signal spectrum automatic detection method, a system and a medium, and discloses the radar pulse signal spectrum automatic detection method, wherein the method comprises the following steps: carrying out digital channelizing on the intermediate frequency complex signal to obtain a sub-channel signal; performing adaptive threshold signal detection to obtain effective signals of the sub-channels; respectively carrying out cross-channel judgment on each pair of adjacent sub-channels, and judging the cross-channel condition of the adjacent sub-channels; carrying out digital channelization on the sub-channels without the cross channel condition, respectively carrying out self-adaptive threshold signal detection, and outputting effective signals of the sub-channels of the next stage; and calculating the center frequency and the bandwidth of the corresponding cross-channel broadband signal for the sub-channel with the cross-channel condition so as to configure the corresponding DDS signal generator and the variable bandwidth filter to perform signal matching and detection on the intermediate frequency complex signal. The invention adopts two-stage channelized detection, realizes automatic updating of noise threshold, and increases the processing of broadband cross-channel signals, thereby being capable of detecting the bandwidth of the self-adaptive broadband radar signals. However, when the prior art is used for detection, the problems of high resource occupation and processing power consumption of the FPGA and difficulty in fast and efficiently detecting the radar signal with the low signal-to-noise ratio complex system still exist.

Disclosure of Invention

The invention aims to provide a radar pulse signal detection method based on spectrum sparse sensing, which is used for solving the technical problems that the FPGA resources are occupied, the processing power consumption is high, and the radar signal of a complex system with low signal-to-noise ratio is difficult to detect rapidly and efficiently in the prior art; the invention also aims to provide a radar pulse signal detection device based on spectrum sparse sensing.

In order to achieve the above purpose, the radar pulse signal detection device based on spectrum sparse sensing adopts the following technical scheme:

the radar pulse signal detection device based on spectrum sparse sensing comprises a spectrum sparse sensing module, wherein the spectrum sparse sensing module comprises a single-bit analog-to-digital converter and a processor, the single-bit analog-to-digital converter is used for carrying out single-bit sampling on an original signal to obtain a discrete sample set, and the processor is used for processing the obtained discrete sample set to obtain signal frequency and signal bandwidth;

the self-adaptive filtering module is electrically connected with the frequency spectrum sparse sensing module to receive the discrete sample set, the signal frequency and the signal bandwidth transmitted by the self-adaptive filtering module, and has the functions of performing down-conversion on the discrete sample set according to the signal frequency and performing low-pass filtering on the down-converted signal according to the signal bandwidth;

and the double-threshold pulse detection module is electrically connected with the dynamic adjustment self-adaptive filtering module to receive the filtered signals transmitted by the dynamic adjustment self-adaptive filtering module and is used for detecting the filtered signals.

Further, the device also comprises a signal acquisition module, which is used for acquiring an original full-frequency band signal, and is electrically connected with the spectrum sparse sensing module so as to transmit the acquired original full-frequency band signal to the spectrum sparse sensing module.

Still further, the device also comprises a signal output module which is electrically connected with the double-threshold pulse detection module to receive the detection result transmitted by the double-threshold pulse detection module and output the detection result.

The radar pulse signal detection method based on spectrum sparse sensing adopts the following technical scheme:

the radar pulse signal detection device based on spectrum sparse sensing comprises the following steps of firstly, collecting an original full-frequency band signal;

secondly, single-bit sampling is carried out on an original full-band signal to obtain a discrete sample set, and the discrete sample set is processed to obtain signal frequency and signal bandwidth;

thirdly, performing down-conversion on the discrete sample set to obtain a zero intermediate frequency signal, and filtering the zero intermediate frequency signal according to the obtained signal bandwidth;

fourth, double threshold pulse detection is performed on the filtered signal.

Further, in the third step, the discrete sample set is down-converted by a mixer, the center frequency of which is adjusted in real time according to the acquired signal frequency.

Further, in the second step, DFT conversion is performed on the acquired discrete sample set to acquire a signal frequency and a signal bandwidth thereof.

Further, in the second step, the discrete sample set is divided into time slices, DFT conversion is performed respectively, then the DFT data of a plurality of time slices within the pulse duration are combined, spectral peak analysis and spectral parameter extraction are performed, the signal frequency and the signal bandwidth are obtained through calculation according to the spectral peak analysis and the spectral peak parameter extraction, and overlapping between adjacent time slices is ensured when the discrete sample set is divided into time slices.

Further, the signal after the third step is

Where x (n) is the input data of the nth sample point, M is the order of the filter, e ^-jωkn Is a mixer, k is the center frequency of the pulse signal, k is a real number, the frequency k of the mixer is consistent with the center frequency k of the pulse signal, W (m) is a prototype low-pass filter, y _k And (n) is the pulse signal after filtering.

The beneficial effects of the invention are as follows: the spectrum sparse sensing module is used for carrying out broadband spectrum sparse real-time sensing on the input radar signal, the symbol bit of the AD data is used for carrying out spectrum sensing, and the spectrum distribution condition of the input is monitored and tracked in real time.

Further, during filtering, according to the central frequency and the bandwidth of the pulse signal obtained by the frequency spectrum monitoring module, a dynamic adjustment self-adaptive filtering module is used for adjusting the central frequency point of the filter, so that tracking filtering of the radar pulse signal is realized. The radar pulse signal detection is realized by using the double-threshold pulse detection module, the signal detection sensitivity is improved, and the detection of the complete pulse signal is ensured. Since the mixing frequency of the mixer module is adjusted in real time for the broadband signal, a single pulse may have multiple mixing frequencies, and compared with a single mixing frequency, the method can further improve the signal-to-noise ratio of the pulse and improve the detection accuracy.

Experiments prove that the invention can detect the 200M broadband linear frequency modulation radar pulse signal with the signal to noise ratio of-10 db and the common radar pulse signal at the same time.

Drawings

FIG. 1 is a block diagram of a radar pulse signal detection device based on spectrum sparse sensing;

FIG. 2 is a detection flow chart of a radar pulse signal detection method based on spectrum sparse sensing according to the present invention;

fig. 3 is a DFT output spectrum of a tone signal.

Fig. 4 is a single bit DFT output spectrum of a single tone signal.

Fig. 5 is a DFT output spectrum of a binaural signal.

Fig. 6 is a single bit DFT output spectrum of a binaural signal.

Fig. 7 is a time domain diagram of simultaneous arrival of a chirp signal superimposed with a conventional signal.

Fig. 8 is a time-frequency diagram of simultaneous arrival of a chirp signal superimposed with a conventional signal.

Fig. 9 is a time domain diagram of a conventional signal after dynamic adaptive filtering.

Fig. 10 is a time domain diagram of a chirped signal after dynamic adaptive filtering.

Fig. 11 is a schematic diagram of dual threshold pulse detection.

Detailed Description

The embodiment of the radar pulse signal detection device based on spectrum sparse sensing comprises:

the specific structure of the radar pulse signal detection device based on spectrum sparse sensing is shown in fig. 1, and the radar pulse signal detection device comprises a signal acquisition module, a spectrum sparse sensing module, a dynamic adjustment self-adaptive filtering module, a double-threshold pulse detection module and an output module.

The signal acquisition module is electrically connected with the spectrum sparse sensing module to transmit the acquired original signals to the spectrum sparse sensing module, the spectrum sparse sensing module comprises a single-bit analog-to-digital converter (ADC) and a processor, the single-bit analog-to-digital converter is used for completing single-bit sampling of the original signals, and the processor is used for processing a discrete sample set obtained by the single-bit sampling to acquire signal frequency and signal bandwidth. The frequency spectrum sparse sensing module is electrically connected with the dynamic adjustment self-adaptive filtering module, on one hand, a discrete sample set obtained by single bit sampling is transmitted to the frequency spectrum sparse sensing module, and on the other hand, the frequency spectrum sparse sensing module transmits the acquired signal frequency and the signal bandwidth to the frequency spectrum sparse sensing module, the dynamic adjustment self-adaptive filtering module performs down-conversion on the discrete sample set obtained by single bit sampling according to the acquired signal center frequency, and performs low-pass filtering on a zero intermediate frequency signal obtained after the down-conversion according to the signal bandwidth. The dynamic adjustment self-adaptive filtering module is electrically connected with the double-threshold pulse detection module and detects the filtered signals. The double-threshold pulse detection module is electrically connected with the output module to transmit detection results to the output module, and the output module is used for outputting the received detection results.

The embodiment of the radar pulse signal detection method based on spectrum sparse sensing comprises the following steps:

a specific flow of a radar pulse signal detection method based on spectrum sparse sensing is shown in fig. 2, which comprises the following steps,

firstly, collecting radar signals;

the acquired radar signal is a complete original full-band signal, which includes noise signals and possibly radar signals.

Calculating a signal frequency spectrum through a frequency spectrum sparse sensing module to obtain a signal frequency and a signal bandwidth;

in the step, a single-bit analog-to-digital converter (ADC) is used for carrying out analog-to-digital conversion on signals input by a signal acquisition module, DFT (discrete Fourier transform) is calculated for signal division time slices in the form of a converted discrete sample set respectively, and certain overlapping exists between adjacent time slices so as to ensure that the DFT processing gain of pulse signals crossing the time slices is not reduced. And merging DFT data of a plurality of time slices within the pulse duration, then carrying out spectral peak analysis and spectral parameter extraction, and calculating to obtain the signal frequency and the signal bandwidth. And the DFT is calculated respectively aiming at the signal dividing time slices in the converted discrete sample set form, so that the signal processing resource can be reduced, and the processing efficiency is improved.

Wherein, the DFT conversion formula of the single-bit sampling signal based on the ADC symbol bit is as follows:

where X (N) is the input data of the nth sample point, N is the number of points of DFT, X (ω) is the output data of DFT, and X (ω) actually reflects the spectral data of the signal.

In the single-bit sampling mode, the value of X (n) is only 0 and 1, so that multiplication operation is not needed when calculating X (omega), and a DFT algorithm can be realized by accumulating the frequency rotation factors e-iomega m by using an accumulator with addition and subtraction control. The DFT algorithm in the invention has the following implementation structure:

the result of DFT output under the condition of 3dB snr of a single tone signal (i.e., a single frequency rf signal) is shown in fig. 3, and the result of DFT output under the condition of 3dB snr of a single tone signal is shown in fig. 4. As can be seen from comparing fig. 3 and fig. 4, the single-bit DFT of the single-tone signal has a small difference from the DFT calculation result of the original signal, and the signal-to-noise ratio of the single-bit DFT is slightly lower.

The DFT output result under the condition of 3dB snr of the dual tone signal (i.e., the radio frequency signal of dual frequencies) is shown in fig. 5, and the single bit DFT output result under the condition of 3dB snr of the dual tone signal is shown in fig. 6. As can be seen by comparing fig. 5 and 6, some spurious spectral components appear in the single bit DFT calculation result compared to the DFT of the original signal. Thus, the single-bit spectrum of the signal has a loss of dynamic range under multitone conditions.

Step three, dynamically adjusting the bandwidth of the adaptive filtering module according to the signal frequency and the signal bandwidth obtained in the step two and realizing filtering;

in the step, down-conversion is carried out on an input signal by utilizing the signal center frequency obtained by spectrum sparse sensing, and low-pass filtering is carried out on a zero intermediate frequency signal according to a signal bandwidth, specifically:

the signals after filtering are:

wherein x (n) has the meaning as above, and is the input data of the nth sampling point, M is the order of the filter, and the value range is generally [256, 384 ]]E-j ωkn is a mixer, k is the center frequency of the pulse signal, k is a real number, the frequency k of the mixer coincides with the center frequency k of the pulse signal, W (m) is a prototype low-pass filter, y _k And (n) is the pulse signal after filtering. Mixing frequency e ^-jωkn And prototype low-pass filter w (n) And the values of the filter coefficients are dynamically adjusted according to the signal frequency and the signal bandwidth obtained by the frequency spectrum sparse sensing module, and self-adaptive filtering is carried out.

When the dynamic adjustment self-adaptive filtering processing is realized along with the change of the frequency parameter, in order to keep the continuity of waveform output, the center frequency of the mixer should be adjusted and updated in real time according to the signal frequency obtained by the frequency spectrum sparse sensing module. Whether the center frequency of the mixer is updated is determined by the amount of change in the signal frequency obtained by the spectrum sparse sensing module, and the center frequency of the mixer is generally updated when the amount of change in the signal frequency exceeds 20 megabytes.

In order to verify the effectiveness of the invention on signal separation by performing adaptive filtering after the signal is subjected to spectrum sparse sensing, the method comprises the following steps:

a set of signals is generated which is synthesized from the simultaneously arriving chirped radar signal and the conventional radar signal, the frequency domain diagram of which is shown in fig. 7 and the time-frequency diagram of which is shown in fig. 8. As can be seen from fig. 7 and 8, since the chirp signal has a wide bandwidth, the spectrum thereof is liable to overlap with that of the conventional signal. But in the time-frequency plane the two signal components do not coincide or only slightly coincide, so that they can be separated by dynamically adjusting the adaptive filter.

The signals are sent to a spectrum sparse sensing module, spectrum components of the two signals are detected by the spectrum sparse sensing module, and two adaptive filters are respectively arranged for filtering separation. For conventional signals, the center frequency of the adaptive filter remains substantially unchanged; for the linear frequency modulation signal, as the frequency of the signal measured by the frequency spectrum sparse sensing module continuously slides, the mixing frequency of the dynamic adjustment adaptive filter correspondingly changes, and time-varying tracking filtering of the frequency modulation signal is realized. After the adaptive filtering process, the time domain signals of the two signal components are obtained as shown in fig. 9 and 10. As can be seen from fig. 9 and 10, after the down-conversion process, the signal becomes a baseband signal; after the adaptive filtering treatment, the signal to noise ratio is obviously improved.

Fourth, performing double-threshold pulse detection on the output signal subjected to the adaptive filtering in the third step

In the step, two thresholds are set by using a double-threshold pulse detection module to ensure the pulse detection integrity, and the detection target is to keep the signal in the set threshold as a detection value, and the specific detection principle is as follows:

the pulse start and end thresholds are:

the pulse duration threshold is:

in the formula

Representing the noise mean value of n sample points (the sample points of each time slice are ordered according to the amplitude, the channel median value after the ordering is the noise in the corresponding channel, and the noise in all channels is averaged to obtain the noise value), and the noise is represented by the noise mean value of n sample points (the sample points of each time slice are ordered according to the amplitude)>

The noise variance of n sample points is represented, and a is a predetermined constant (the value of a is generally determined by the false alarm rate, and the default value is set to 4).

A schematic diagram of the principle of using dual threshold pulse detection is shown in fig. 11, where the upper edge line represents the pulse start and end thresholds, the lower edge line represents the pulse duration threshold, the middle line represents the threshold average, and the signal between the pulse start and end thresholds and the pulse duration threshold (i.e., the signal set in the middle) is the signal that remains after detection. The bandwidth, frequency and pulse width of the pulse signal can be measured more accurately by using double-threshold pulse detection, and the sensitivity of pulse signal detection is greatly improved.

And fifthly, outputting a detection result.

While the invention has been described in detail in the foregoing general description and specific examples, it will be apparent to those skilled in the art that modifications and improvements can be made thereto. Accordingly, such modifications or improvements may be made without departing from the spirit of the invention and are intended to be within the scope of the invention as claimed.

Claims

1. Radar pulse signal detection device based on frequency spectrum sparse perception, its characterized in that: the device comprises a spectrum sparse sensing module, wherein the spectrum sparse sensing module comprises a single-bit analog-to-digital converter and a processor, the single-bit analog-to-digital converter is used for carrying out single-bit sampling on an original signal to obtain a discrete sample set, and the processor is used for processing the obtained discrete sample set to obtain signal frequency and signal bandwidth;

2. The radar pulse signal detection apparatus based on spectrum sparse sensing according to claim 1, wherein: the system also comprises a signal acquisition module, a spectrum sparse sensing module and a spectrum sparse sensing module, wherein the signal acquisition module is used for acquiring an original full-frequency band signal and is electrically connected with the spectrum sparse sensing module so as to transmit the acquired original full-frequency band signal to the spectrum sparse sensing module.

3. The radar pulse signal detection apparatus based on spectrum sparse sensing according to claim 1, wherein: the signal output module is electrically connected with the double-threshold pulse detection module to receive the detection result transmitted by the double-threshold pulse detection module and output the detection result.

4. Radar pulse signal detection device based on frequency spectrum sparse perception, its characterized in that: the method comprises the following steps that firstly, an original full-frequency band signal is collected;

fourth, double threshold pulse detection is performed on the filtered signal.

5. The radar pulse signal detection method based on spectrum sparse sensing according to claim 4, wherein: in the third step, the discrete sample set is down-converted by a mixer, the center frequency of which is adjusted in real time according to the acquired signal frequency.

6. The radar pulse signal detection method based on spectrum sparse sensing according to claim 4, wherein: in the second step, a DFT transform is performed on the acquired set of discrete samples to find its signal frequency and signal bandwidth.

7. The radar pulse signal detection method based on spectrum sparse sensing according to claim 6, wherein: in the second step, DFT conversion is performed respectively for the discrete sample set dividing time slices, then, spectral peak analysis and spectral parameter extraction are performed after DFT data of a plurality of time slices within pulse duration are combined, and the signal frequency and the signal bandwidth are obtained by calculation according to the spectral peak analysis and the spectral parameter extraction, and overlapping between adjacent time slices is ensured when the discrete sample set dividing time slices are performed.

8. The radar pulse signal detection method based on spectrum sparse sensing according to claim 5, wherein: the signal after the third step is

Wherein x (n) is the thThe input data of n sampling points, M is the order of the filter, e-jωkn is a mixer, k is the center frequency of the pulse signal, k is a real number, the frequency k of the mixer is consistent with the center frequency k of the pulse signal, W (M) is a prototype low-pass filter, y _k And (n) is the pulse signal after filtering. />