Broadband pulse signal acquisition and analysis system

By designing a broadband pulse signal acquisition and analysis system, using a combination of multiple detection algorithms and modules, the anti-interference and real-time problems of pulse signal processing in high-frequency and high-speed environments are solved, and efficient and reliable pulse signal detection and storage are achieved.

CN118914677APending Publication Date: 2024-11-08CHANGSHA ZHIDIAN INFORMATION TECHNOLOGY CO LTD

Patent Information

Application Number
CN202411086275.1
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-08
Publication Date
2024-11-08

AI Technical Summary

Technical Problem

The existing pulse signal processing technology has poor anti-interference ability, large energy loss, narrow bandwidth range in high frequency, high speed and complex environments, and insufficient real-time and data security.

Method used

Design a broadband pulse signal acquisition and analysis system, including input and RF conversion module, broadband signal acquisition module, signal processing and data analysis module, signal high-speed storage module and human-computer interaction module, and receive broadband RF signals in real time, downconversion, and digitization. Sampling, real-time pulse detection and storage adopts GO-CFAR and OS-CFAR detection methods, instantaneous autocorrelation phase jump detection algorithms and box differential detection filters to achieve efficient pulse signal processing and storage.

Benefits of technology

It improves the anti-interference ability of high-frequency signal processing and the real-time nature of data storage, enhances the detection accuracy and storage management capabilities of pulse signals, and meets the needs of efficient and reliable pulse signal processing and management.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118914677A_ABST
    Figure CN118914677A_ABST
Patent Text Reader

Abstract

The invention relates to a broadband pulse signal acquisition and analysis system, which comprises an input and radio frequency conversion module for receiving a broadband radio frequency signal in real time and performing down-conversion on the radio frequency broadband signal to adjust the radio frequency broadband signal into an intermediate frequency signal, a broadband signal acquisition module for performing digital sampling on the intermediate frequency signal and converting the intermediate frequency signal into corresponding baseband signal IQ data, and a signal processing module for processing the baseband signal IQ data. The signal processing and data analysis module is used for carrying out real-time pulse detection and data afterglow statistics on baseband signal IQ data according to data analysis related parameters; the signal high-speed storage module is used for carrying out real-time storage on detected pulse data and complete pulse description words according to storage related information; and the man-machine interaction module is used for acquiring data analysis related parameters and storage related information through the interactive interface, respectively sending the data analysis related parameters and the storage related information to the corresponding modules, and displaying pulse detection results through the interactive interface. By adopting the system, the problems of high-frequency signal processing, digital sampling and processing, pulse signal processing, real-time storage and management, user customization requirements and the like can be solved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to the technical field of pulse signal analysis, and in particular to a broadband pulse signal acquisition and analysis system. Background Art

[0002] The use of pulse signals has broad application prospects and important significance in the fields of medical care, communication engineering, aerospace, military, etc. Therefore, pulse signal acquisition and analysis technology has received more and more attention and research from researchers at home and abroad.

[0003] However, although the application of pulse signals has made significant progress, there are still some technical problems that need to be solved. On the one hand, the traditional pulse signal processing method has certain limitations in high frequency, high speed and complex environment, such as weak anti-interference ability, large energy loss, narrow bandwidth, etc. On the other hand, the existing pulse signal transmission technology still has some shortcomings in real-time, reliability and data security. Summary of the invention

[0004] Based on this, it is necessary to provide a broadband pulse signal acquisition and analysis system that can solve the problems of poor real-time performance of pulse signal transmission technology and weak anti-interference ability of pulse signal processing in complex environments in response to the above technical problems.

[0005] A broadband pulse signal acquisition and analysis system comprises: an input and radio frequency conversion module, a broadband signal acquisition module, a signal processing and data analysis module, a signal high-speed storage module connected in sequence, and a human-computer interaction module connected to the signal processing and data analysis module and the signal high-speed storage module respectively;

[0006] The input and RF conversion module is used to receive broadband RF signals in real time, down-convert the RF broadband signals into intermediate frequency signals, and send the intermediate frequency signals to the broadband signal acquisition module;

[0007] The broadband signal acquisition module is used to digitally sample the intermediate frequency signal, convert it into corresponding baseband signal IQ data, and send the baseband signal IQ data to the signal processing and data analysis module;

[0008] The signal processing and data analysis module is used to perform real-time pulse detection and data afterglow statistics on the baseband signal IQ data according to data analysis related parameters to obtain pulse detection results, wherein when the baseband signal IQ data is subjected to real-time pulse detection, signal detection is performed on the baseband signal IQ data, and if a signal exists, time domain parameters and frequency domain parameters are detected;

[0009] The signal high-speed storage module is used to store the detected pulse signal in real time according to the storage related information;

[0010] The human-computer interaction module is used to obtain the data analysis related parameters and storage related information through the interactive interface, and send them to the signal processing and data analysis module and the signal high-speed storage module respectively, and also display the pulse detection results through the interactive interface.

[0011] In one embodiment, in the signal processing and data analysis module, when real-time pulse detection and digital afterglow statistics are performed on the baseband signal IQ data according to data analysis related parameters to obtain pulse detection results:

[0012] Perform signal detection on the baseband signal IQ data, and if a signal exists, perform time domain detection to obtain time domain parameters, or perform frequency domain detection using an instantaneous autocorrelation phase jump detection algorithm to obtain frequency domain parameters;

[0013] Perform digital afterglow statistics according to the baseband signal IQ data to obtain digital afterglow statistics results;

[0014] The time domain parameters, frequency domain parameters and digital afterglow statistics are used as the pulse detection results.

[0015] In one embodiment, when performing signal detection on the baseband signal IQ data:

[0016] The baseband signal IQ data includes multiple frames of data sorted in time, and signal detection is performed on each frame of data in the baseband signal IQ data in turn;

[0017] The GO constant false alarm rate detection method is used to perform pulse detection on the data after fast Fourier transformation. If a pulse signal is detected, a time domain or frequency domain detection is performed;

[0018] If no pulse signal is detected, it is determined whether the number of frames with pulse edges before the current frame data exceeds half. If it exceeds half, the OS constant false alarm rate detection method is used to perform secondary pulse detection on the current frame data.

[0019] In one embodiment, when performing time domain detection on the baseband signal IQ data of the existing signal:

[0020] After performing short-time Fourier transform on the baseband signal IQ data, the MN algorithm is used to perform coarse pulse edge detection. If a pulse edge is detected, a box differential detection filter is used to perform fine detection at the location of the pulse edge.

[0021] In one embodiment, the frequency range of the real-time received broadband radio frequency signal is 1-18 GHZ, the center frequency of the intermediate frequency signal is 1.8 GHz, and the effective bandwidth is 1 GHz.

[0022] In one embodiment, the input and RF conversion module includes a plurality of pre-selection filter banks and a broadband down-converter;

[0023] The plurality of pre-selection filter groups filter the broadband radio frequency signals received in real time;

[0024] The broadband down-converter uses its internally integrated low phase noise local oscillator to convert the filtered broadband RF signal into an intermediate frequency signal, and performs amplitude adjustment and anti-aliasing filtering on the intermediate frequency signal before sending it to the broadband signal acquisition module.

[0025] In one embodiment, the high-speed data storage module uses an embedded solid-state disk array.

[0026] In one embodiment, the signal processing and data analysis module includes an FPGA unit and a GPU unit;

[0027] In the FPGA unit, signal detection is performed on the baseband signal IQ data. If a signal exists, time domain detection is performed to obtain time domain parameters, and the pulse data is transmitted to the GPU unit through the PCle bus to perform digital afterglow statistics to obtain digital afterglow statistics results;

[0028] In the GPU unit, frequency domain detection is performed using an instantaneous autocorrelation-based phase jump detection algorithm to obtain frequency domain parameters.

[0029] The above-mentioned broadband pulse signal acquisition and analysis system includes an input and RF conversion module for receiving broadband RF signals in real time and down-converting the RF broadband signals to intermediate frequency signals, a broadband signal acquisition module for digitally sampling the intermediate frequency signals and converting them into corresponding baseband signal IQ data, a signal processing and data analysis module for performing real-time pulse detection and data afterglow statistics on the baseband signal IQ data according to data analysis related parameters to obtain pulse detection results, wherein, when performing signal detection on the baseband signal IQ data, if there is a signal, the time domain parameters and frequency domain parameters are detected, and the signal high-speed storage module for real-time storage of the detected pulse signal according to the storage related information, and a human-computer interaction module for obtaining data analysis related parameters and storage related information through an interactive interface, and sending them to the signal processing and data analysis module and the signal high-speed storage module respectively, and also displaying the pulse detection results through the interactive interface. The use of this system can solve the problems of high-frequency signal processing, digital sampling and processing, pulse signal processing, real-time storage and management, and user customization needs, and provides an efficient and reliable solution for applications in related fields. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 A structural block diagram of a broadband pulse signal acquisition and analysis system in one embodiment;

[0031] Figure 2 A schematic diagram of the working principle of a broadband down converter in one embodiment;

[0032] Figure 3 A schematic diagram of the architecture of an FMC high-speed ADC acquisition daughter card in one embodiment;

[0033] Figure 4 A schematic diagram of a flow chart of a signal sampling function in one embodiment;

[0034] Figure 5 A schematic diagram of the relationship between the bandpass sampling frequency in one embodiment;

[0035] Figure 6 A schematic diagram of a channelization scheme of a digital receiver in one embodiment;

[0036] Figure 7 A schematic diagram of a process of performing real-time pulse detection on baseband signal IQ data in one embodiment;

[0037] Figure 8 It is a schematic diagram of a process of performing pulse detection on baseband signal IQ data in one embodiment;

[0038] Fig. 9 A flow chart of signal modulation type identification based on an instantaneous autocorrelation phase jump detection algorithm in one embodiment;

[0039] Fig.10 A schematic diagram of digital afterglow statistics in one embodiment;

[0040] Fig.11 is a structural block diagram of a signal processing and data analysis module in one embodiment;

[0041] Fig.12 is a schematic diagram of the architecture of a signal processing and data analysis module in one embodiment;

[0042] Fig.13 is a system block diagram of a high-speed solid-state memory card in one embodiment;

[0043] Fig.14 It is a block diagram of the Direct Stream technology structure based on the PCIe3.0 bus in one embodiment;

[0044] Fig.15 The overall architecture diagram of a broadband pulse signal acquisition and analysis system in one embodiment. DETAILED DESCRIPTION

[0045] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application 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 application and are not used to limit the present application.

[0046] In view of the problems in the prior art such as weak anti-interference ability, large capacity loss, narrow bandwidth range, etc. when processing pulse signals at high frequency, high rate and complex environment, in this application, Figure 1 As shown, a broadband pulse signal acquisition and analysis system 10 is provided, which includes an input and RF conversion module 101, a broadband signal acquisition module 102, a signal processing and data analysis module 103, a signal high-speed storage module 104, and a human-computer interaction module 105 respectively connected to the signal processing and data analysis module 103 and the signal high-speed storage module 104.

[0047] Among them, the input and RF conversion module 101 is used to receive broadband RF signals in real time, down-convert the RF broadband signals into intermediate frequency signals, and send the intermediate frequency signals to the broadband signal acquisition module 102. The broadband signal acquisition module 102 digitally samples the intermediate frequency signals, converts them into corresponding baseband signal IQ data, and sends the baseband signal IQ data to the signal processing and data analysis module 103. The signal processing and data analysis module 103 performs real-time pulse detection and data afterglow statistics on the baseband signal IQ data according to the data analysis related parameters to obtain the pulse detection result, wherein when the baseband signal IQ data is subjected to real-time pulse detection, the baseband signal IQ data is subjected to signal detection, and if there is a signal, the frequency domain parameters and time domain parameters are detected. The signal high-speed storage module 104 stores the detected pulse information in real time according to the storage related information. The human-computer interaction module 105 obtains the data analysis related parameters and the storage related information through the interactive interface, and sends them to the signal processing and data analysis module 103 and the signal high-speed storage module 104 respectively, and also displays the pulse detection result through the interactive interface.

[0048] Specifically, the output end of the input and RF conversion module 101 is connected to the input end of the signal acquisition module 102, the output end of the signal acquisition module 102 is connected to the input end of the signal processing and data analysis module 103, the output end of the signal processing and data analysis module 103 is connected to the input end of the signal high-speed storage module 104 and the human-computer interaction module 105, and the human-computer interaction module 105 interacts with the high-speed storage module 104 and the signal processing and data analysis module 103 respectively.

[0049] In this embodiment, the input and RF conversion module 101 implements the functions of limiting, attenuating, filtering, frequency conversion and amplifying the input signal, ie, the broadband RF signal received in real time.

[0050] Specifically, the input and RF conversion module 101 includes multiple groups of pre-selection filter groups and broadband down-converters. The multiple groups of pre-selection filter groups first filter the broadband RF signals received in real time, and then the broadband down-converter uses the internally integrated low phase noise local oscillator to convert the filtered broadband RF signals into intermediate frequency signals, and then performs amplitude adjustment and anti-aliasing filtering on the intermediate frequency signals before sending them to the broadband signal acquisition module 102.

[0051] Furthermore, multiple groups of pre-selected filter groups are integrated at the input end of the input and RF conversion module 101, which has the characteristics of large dynamic range, strong anti-interference ability and high anti-burning power. At the same time, the input and RF conversion module 101 is also integrated with a broadband down-converter with down-conversion and gain adjustment functions, which down-converts the RF signal to an intermediate frequency signal, adjusts the amplitude of the output signal, and provides the intermediate frequency signal to the broadband down-converter of the broadband signal acquisition module 102. The broadband down-converter has a low phase noise local oscillator source integrated inside, has the characteristics of fast frequency conversion time, and can achieve fast full-band scanning. .

[0052] like Figure 2 The figure shows the working principle of a broadband downconverter. In a broadband downconverter, the input data is mixed with a local oscillator to generate an intermediate frequency signal. The frequency of the intermediate frequency signal is adjusted by adjusting the frequency of the local oscillator. At the same time, the filter removes the redundant frequency components to ensure the purity of the output intermediate frequency signal. The downconversion process is usually accompanied by power changes, and the power control circuit is required to maintain a stable power level. Finally, the adjusted intermediate frequency signal is sent to the next stage of processing or transmission to achieve effective adjustment and adaptation of the broadband RF signal.

[0053] In this embodiment, the frequency range of the broadband radio frequency signal received in real time is 1-18 GHZ, and a single frequency conversion scheme is adopted to convert the frequency to a fixed intermediate frequency (IF) frequency of 1.8 GHz, and the IF bandwidth is 1 GHz.

[0054] Specifically, in the broadband downconverter, it mainly consists of three parts: frequency conversion part, local oscillator part and reference part. The downconversion process is as follows: first, the 1G-18GHz RF signal is divided into multiple segments, each segment is first mixed with the adjustable local oscillator to obtain the high-IF signal, and then mixed with the second-stage fixed local oscillator to obtain the 1.8G±500MHz intermediate frequency signal. The reference part provides a 100MHz reference and controls the switching function of the internal and external references.

[0055] In the signal input and RF conversion module 101, the broadband down-converter integrates multiple groups of pre-selected filter groups at the input end, which has the characteristics of large dynamic range, strong anti-interference ability, and high anti-burnout power. Its internal integrated low phase noise local oscillator has the characteristics of short frequency conversion time, which can realize fast full-band scanning. Placing each filter in front of the corresponding mixer can realize the frequency conversion function without image interference, and finally convert the input signal to IF. This multi-order mixing scheme specifically uses broadband filters, which can perform image-free conversion over the entire acquisition bandwidth and achieve repeatable specified amplitude flatness and phase linearity.

[0056] In the broadband downconverter, a superheterodyne system is used to convert the RF input signal to a fixed intermediate frequency, which is convenient for subsequent digital sampling and filtering. At the same time, the superheterodyne design can improve the frequency test range of the equipment and increase the sensitivity of the equipment. The local oscillator source of the first mixing module uses frequency synthesis technology to increase the channel selectivity of the frequency converter. Through two frequency conversion processes, it can not only realize the flexible configuration of the receiving channel, but also obtain a better image frequency suppression ratio.

[0057] In this embodiment, the broadband signal acquisition module 102 is used to perform high-speed ADC and digital down-conversion DDC conversion on the intermediate frequency signal after the RF conversion processing, and output the baseband signal IQ data to the signal processing and data analysis module.

[0058] Specifically, the broadband signal acquisition module 102 has a standard FMC (VITA 57.1) interface, and its high-speed ADC chip can adopt a chip whose main functions and performance parameters meet the following requirements: ADC sampling rate: up to 6GSps, resolution: 12 bits, input power: ≤3dBm, analog input bandwidth: 6GHz, trigger: support external trigger input, instantaneous dynamic range: 60dBc (typical value).

[0059] In one embodiment, the ADC chip in the broadband signal acquisition module 102 may be a chip of model: ADC12D3200RF.

[0060] In one embodiment, the broadband signal acquisition module 102 is implemented using an FMC high-speed ADC acquisition daughter card, and its functional block diagram is as follows: Figure 3 shown.

[0061] Further, according to the bandpass acquisition principle, the signal sampling design scheme is as follows: the RF signal undergoes frequency conversion and gain adjustment through a broadband downconverter, and the output intermediate frequency signal needs to undergo anti-aliasing filtering before entering the broadband signal acquisition module. The digital receiver in the broadband signal acquisition module 102 adopts a bandpass sampling scheme. The necessary and sufficient conditions for the sampling rate of the bandpass signal to reconstruct the signal without distortion are:

[0062]

[0063] In the above formula, f0 represents the center frequency, B represents the bandwidth, and f s Indicates the sampling rate, m max =|f H / B|, m represents the Quesnaisd interval.

[0064] Under the above conditions, if the following equation holds, the frequency mixing engineering can be easily realized.

[0065]

[0066] In order to achieve direct sampling of 1000MHz bandwidth, a sampling rate of 2400MSps is used to sample the 1.8GHz intermediate frequency signal without distortion. At the same time, considering the out-of-band spectrum components of the signal, the sampling frequency should be appropriately greater than twice the signal bandwidth so that spectrum aliasing will not occur after sampling. The selection of bandwidth margin should be considered comprehensively. A large margin can reduce spectrum aliasing and reduce the order of the orthogonal demodulation filter, but it also increases the sampling and processing clock frequency, increasing data transmission and storage pressure.

[0067] For the intermediate frequency sampling part, when the center frequency is 1800MHz and the bandwidth is 1000MHz, a feasible solution is 2400MSps sampling rate (m=3). The signal flow is as follows: Figure 4 shown.

[0068] like Figure 5 As shown in FIG. 1 , it is the principle of bandpass sampling. It can be clearly seen that when the sampling rate Fs=2400MSps, the frequency band of the broadband signal of the intermediate frequency input is exactly located in the third Nyquist interval of the bandpass sampling.

[0069] In this embodiment, the indicators used for the wideband signal are: sampling rate: 2400MSps, number of ADC channels: 1, ADC bit width: 12 bits, signal input power: -60 to 0dBm, and spurious-free dynamic range: 60dBc (typical value).

[0070] In this embodiment, a channelized receiving and processing scheme is adopted. The sampling rate of the digital receiver ADC is Fs=2400MSps. After the first stage of digital down-conversion, the output IQ sampling rate is Fs=1200MSps, and the effective signal bandwidth is 1GHz. Through digital filtering and extraction, the signal-to-noise ratio is further improved while reducing the processing bandwidth. The functional block diagram is shown in Figure 6. The original signal recording scheme has a signal sampling frequency of Fs=2400MSps, a sampling bit width of 12 bits, and is stored as 16 bits. After channelized receiving and processing, an IQ data stream of 1200MSps is obtained by filtering and down-sampling. The data volume statistics after considering different extraction multiples are shown in the following table.

[0071] Table 1 Summary of data rates under different bandwidth conditions

[0072]

[0073]

[0074] In this embodiment, in the signal processing and data analysis module 103, when the baseband signal IQ data is subjected to real-time pulse detection and digital afterglow statistics according to the analysis of relevant parameters to obtain the pulse detection result: firstly, the baseband signal IQ data is subjected to signal detection, and if there is a signal, time domain detection is performed to obtain time domain parameters, or frequency domain detection is performed using an instantaneous autocorrelation phase jump detection algorithm to obtain frequency domain parameters. At the same time, digital afterglow statistics are performed according to the baseband signal IQ data to obtain digital afterglow statistics results, and finally, the time domain parameters, frequency domain parameters and digital afterglow statistics results are used as the pulse detection results.

[0075] In the embodiment, during the pulse signal detection process, constant false alarm rate (CFAR) detection is used, which is a method dedicated to providing predictable detection and false alarm in an actual interference environment, also known as "adaptive threshold detection" or "automatic detection". The goal of CFAR detection is to control the false alarm probability at a stable level, so the intensity of noise and interference should be estimated in real time during the detection process.

[0076] Furthermore, we selected the GO-CFAR detection method, i.e., the GO constant false alarm rate detection method, which has good clutter suppression capability, and the OS-CFAR detection method, i.e., the OS constant false alarm rate detection method, which has good effect in suppressing the performance degradation caused by the shielding effect. These two CFAR detection methods are combined to effectively detect pulse signals.

[0077] like Figure 8 As shown, when signal detection is performed on the baseband signal IQ data, the baseband signal IQ data includes multiple frames of data sorted by time, and signal detection is performed on each frame of the baseband signal IQ data in turn. The GO constant false alarm rate detection method is first used to perform pulse detection on the data after the fast Fourier transform. If a pulse signal is detected, time domain or frequency domain detection is performed. If no pulse signal is detected, it is determined whether the number of frames with pulse edges before the current frame data exceeds half. If it exceeds half, the OS constant false alarm rate detection method is used to perform secondary pulse detection on the current frame data.

[0078] Specifically, for the k-th frame data, perform FFT transformation on it. After the transformation, first use GO-CFAR to detect whether there is a signal in the current frame. If not detected, determine whether the number of frames with signals in the previous p frames exceeds half. If it exceeds half, use OS-CFAR to perform secondary detection on the current frame signal. In this way, when performing rough detection of pulse signals, it can effectively prevent missed detection caused by signal imbalance or short-term shielding, and has a certain effect on reducing pulse breakage when the signal-to-noise ratio is not high.

[0079] like Figure 7 As shown, when performing time domain detection on the baseband signal IQ data with a signal, after performing short-time Fourier transform on the baseband signal IQ data, the MN algorithm is first used to perform coarse pulse edge detection. If a pulse edge is detected, a box differential detection filter is used at the position of the pulse edge for fine detection, and its accuracy can reach ≤0.1us.

[0080] Specifically, when detecting time domain parameters, the mn method based on short-time Fourier transform (STFT) combined with STFT and mn method can effectively detect the pulse edge when the signal-to-noise ratio is not high. The specific process is as follows:

[0081] (1) A rectangular window of length Nwin is added to the real sampled signal s to be detected, and a short-time Fourier transform is performed. The frequency value corresponding to the spectrum peak is selected to represent the frequency of the window;

[0082] (2) Slide the rectangular window from the head of the signal s once, with the size of each sliding window being Nstep, perform short-time Fourier transform on the signal in the window, and select the frequency value as in (1) to obtain a set of frequencies f1, f2, ..., fi, ..., fM. Assuming that the length of the signal s is LEN, then:

[0083] M=fix[(LEN-N win ) / N step ]+1

[0084] In the above formula, fix is ​​the rounding function.

[0085] (3) Setting a signal power threshold value fT, comparing f1, f2, ..., fi, ..., fM with fT, and setting the value to 1 if the value is greater than fT, and setting the value to 0 if the value is not; or using the aforementioned CFAR method to determine whether there is a signal in each window, and setting the value to 1 if the value is greater than fT, and setting the value to 0 if the value is not;

[0086] (4) Since there is noise in the signal s and the noise is random, it may also be within the threshold range and be mistaken for a signal. In order to reduce the probability of false alarm, this paper uses the mn method to determine the existence of signal pulses: that is, if n of the m consecutive data are 1, it is considered that there is a signal, and its starting point TOAi is the moment when k 1s appear for the first time among these m points. Otherwise, it is considered that there is no signal.

[0087] When the starting point of the signal is detected, if more than n 0s appear in the continuous m signals, the signal is considered to have ended, and its end point TOEi is the moment when k 0s appear continuously for the first time in these m points, and then the next signal detection is continued. The pulse width (PW) and pulse repetition period (PRI) of the corresponding pulse are:

[0088] PW i=TOE i -TOA i

[0089] PRI i =TOA i+1 -TOA i

[0090] Since the STFT-based mn method detects signals in the frequency domain, its measurement accuracy of time domain parameters is related to the window size Nstep, and its time domain resolution is:

[0091] δ t =N step / 2Fs

[0092] In the above formula, Fs is the sampling frequency. For example, if the sampling frequency Fs is 2.4G, the length of the short-time Fourier transform Nwin is 1024 points, and the sliding window length Nstep is also 1024, that is, the data is not repeated, the time domain resolution is about 214ns; if Nstep is 100, the time domain resolution can reach 10.4ns. However, it should be noted that the amount of algorithm calculation will also increase proportionally at this time. It takes 10 FFTs to process 1024 points, and the system load increases sharply.

[0093] Specifically, when performing precise detection of pulse signals, a box differential filter, namely a DOB filter, is used. The DOB filter is one of the commonly used filters in digital image processing. The detection of pulse signals is generally based on the intermediate frequency digitized signal, and the leading and trailing edges of the pulse are detected from the noisy signal. There will also be discontinuity between the pulse signal part and the pure noise part between the pulses, so the idea of ​​digital image edge detection can be used to detect the pulse.

[0094] For causal systems, the DOB filter h(k) of size N (order 2N+1) is:

[0095]

[0096] Assume the intermediate frequency sampling signal is:

[0097] x(k)=s(k)+n(k)

[0098] In the above formula, s(k) is the radar signal and n(k) is the noise. The signal length is M, the filter h(k) order is 2N+1, and the output signal y(k) length is M+2N. Taking the absolute value of the intermediate frequency sampling signal as input, the output signal is:

[0099]

[0100] Let the sampling duration be \(t\), the time of arrival (TOA) of the pulse signal be \(t_1\), and the time of end (TOE) of the pulse be \(t_2\). For the convenience of analysis, assume that the sampled data contains only one pulse, and the number of sampling points of the pulse signal and the noise parts before and after the pulse are not less than the filter order, that is, \(t_1\), \(t_2 - t_1\), and \(t - t_2\) are all not less than \((2N + 1)T\). s 。

[0101] Analysis shows that:

[0102] (1) When \(kT\) s \(< t_1\), \(y(k)=0\);

[0103] (2) When \((k - N)T\) s \(< t_1\leq kT\) s , \(y(k)>0\) and is non - decreasing monotonically with the increase of \(k\);

[0104] (3) When \((k - 2n)T\) s \(< t_1\leq(k - n)T\) s , \(y(k)>0\) and is non - increasing monotonically with the increase of \(k\);

[0105] (4) When \(t_1<(k - 2N)T\) s \(\leq kT\) s \(\leq t_2\), \(y(k)=0\);

[0106] (5) When \((k - N)T\) s \(< t_2\leq kT\) s , \(y(k)<0\) and is non - increasing monotonically with the increase of \(k\);

[0107] (6) When \((k - 2N)T\) s \(< t_2\leq(k - N)T\) s , \(y(k)<0\) and is non - decreasing monotonically with the increase of \(k\);

[0108] (7) When \(t_2<(k - 2N)T\) s , \(y(k)=0\).

[0109] Thus, when \(k = [t_1 / T\) s +N ([] represents the floor operation), denoted as \(K\) max , \(y(k)\) reaches the maximum value; when \(k = [t_2 / T\) s +N, denoted as \(K\) min , \(y(k)\) reaches the minimum value. From this, the pulse can be detected and the TOA and TOE of the pulse can be accurately calculated:

[0110]

[0111] Pulse width:

[0112] W p=t2-t1=(k min -k max )T s

[0113] In the above formula, the measurement accuracy of time domain parameters is T s .

[0114] Next, the pulse data with data format are subjected to frequency domain detection by instantaneous autocorrelation and digital afterglow statistics.

[0115] Among them, frequency domain detection by instantaneous autocorrelation is a commonly used method in the field of signal processing, especially in spectrum analysis and frequency domain feature extraction. The basic steps include: dividing the baseband analysis signal obtained by preprocessing into frames according to a certain time window to obtain a series of sub-signal fragments. These sub-signal fragments will be used as the input of instantaneous autocorrelation. Perform instantaneous autocorrelation with an interval of m and m-point smoothing accumulation processing; by calculating the instantaneous autocorrelation, the correlation information of each sub-signal fragment in the frequency domain can be obtained. Use the calculated instantaneous autocorrelation results for frequency domain analysis. Spectral analysis based on Fourier transform, adaptive filter, etc. Through frequency domain analysis, the characteristic information of the signal in the frequency domain can be extracted. According to the results of frequency domain analysis, frequency domain detection is performed.

[0116] In order to improve the anti-noise performance, the instantaneous autocorrelation is improved. The instantaneous autocorrelation value is smoothed by an m-point sliding window and then artg is obtained, that is:

[0117]

[0118] In the above formula, the coefficient 1 / m has no effect on the value of artg[Rs(n,m)] and can be discarded. When the nth point and the n+mth point are located before and after the phase jump point, Rs(n,m) is independent of the position of n. Due to multiple superpositions, Rs(n,m) reduces the requirements for the signal-to-noise ratio. After obtaining the phase difference from Rs(n,m), the modulus is taken, and then the phase difference is smoothed by a p-point sliding window to obtain the smoothed phase difference. The number of points p should be selected reasonably. If p is too small, it will affect the smoothing effect and make the phase difference curve have too many burrs; if p is too large, the phase jump peak will be widened and the phase jump amplitude will be weakened.

[0119] The comparison shows that the p value should not be higher than m / 2. This algorithm can make the frequency domain measurement accuracy ≤1MHz. Based on the instantaneous autocorrelation phase jump detection algorithm, the debugging type can be identified, such as Fig. 9 shown.

[0120] In this embodiment, digital afterglow statistics are performed while pulse detection is implemented, and a digital afterglow diagram, a time domain diagram, and a spectrum diagram can be drawn, and the pulse sorting results can be displayed in the form of a bar graph and a list.

[0121] Digital afterglow statistics converts the signal from the time domain to the frequency domain through continuous real-time discrete Fourier transform, then writes the transformed spectrum into the bitmap database at full speed for accumulation, and then converts it into an intuitive and dynamic digital spectrum afterglow graph. Fig.10 As shown in the figure, the essence of digital persistence statistics is to integrate all signal spectra within a period of time into a spectrum graph for display. The horizontal axis of the graph represents frequency, the vertical axis represents amplitude, and the color is used to represent the number of hits of each cell within a certain period of time. In the bitmap library database, the spectrum of each frame is converted into a two-dimensional matrix for representation. The row represents frequency, the column represents spectrum amplitude, and the value of each cell formed by these rows and columns represents the number of spectrum hits entering the grid. The energy distribution of the signal at different frequencies can be obtained to analyze the spectrum characteristics of the signal.

[0122] In this embodiment, the signal processing and data analysis module includes an FPGA unit and a GPU unit. In the FPGA unit, the baseband signal IQ data is subjected to real-time pulse detection to obtain a pulse signal with a data format, and the pulse signal with a data format is transmitted to the GPU unit through a PCle bus. In the GPU unit, according to the pulse signal with a data format, a complete pulse description word is extracted using an instantaneous autocorrelation phase jump detection algorithm, and digital afterglow statistics are performed to obtain digital afterglow statistics results.

[0123] like Fig.11 As shown, the signal processing and data analysis module is composed of an FPGA carrier board and a GPU processing card. Specifically, it is composed of a 6U VPX FPGA carrier board and a GPU processing card.

[0124] In one embodiment, a block diagram of a carrier system of a signal processing and data analysis module is provided, such as Fig.12 As shown in the figure. The carrier board is composed of two FPGA processing architectures, with a Virtex-7 series XC7VX690T and a Zynq series XC7Z030 as standard configurations, and provides abundant logic resources and IO interfaces. The XC7VX690T supports two fully defined FMC+ interfaces onboard, which can be installed with standard FMC+ (VITA 57.1) interface high-speed acquisition and playback cards, and one FMC interface for low-speed interface expansion.

[0125] First, when the acquisition subcard, i.e., the broadband signal acquisition module, is inserted, its ADC (analog-to-digital converter) is responsible for sampling the input signal and converting it into baseband signal IQ data through DDC. The baseband signal IQ data is then transmitted to the FPGA through the FMC+ connector for time domain (frequency domain) signal detection, and outputs formatted pulse data. When the processed data rate is relatively high, it can be cached in the FPGA's peripheral DDR3 and processed in real time, or connected to other components through the FPGA's internal bus for further processing. The pulse data with data format is then transmitted to the host computer's high-performance GPU card through the PCle bus for frequency domain (time domain) parameter measurement, i.e., the instantaneous autocorrelation phase jump detection algorithm and digital afterglow statistics are performed.

[0126] Adopting CPU+GPU heterogeneous processing architecture, the data processing rate can reach hundreds of MB / s. Adopting multi-core CPU or dedicated high-performance GPU processing card, the processing rate can reach hundreds of MB / s, with powerful pulse description word extraction function, real-time pulse maximum data rate: ≥2*10 5 Pulses / second.

[0127] On the other hand, if a playback daughter card is inserted, which refers to a revisit module that can be added to the broadband pulse signal acquisition and analysis system, the main controller transmits the data in the high-speed data storage module to the V7 FPGA through the PCIE interface. In the FPGA, relevant signal processing operations are performed on the incoming data, and the processed data is transmitted to the playback daughter card through the FMC+ connector. The two FMC+ slots can adapt to standard FMC+ or FMC acquisition and playback daughter cards in different application scenarios. Signal conversion is achieved through ADC and DAC conversion, FPGA data processing and PCIE interface data transmission.

[0128] In this embodiment, the high-speed data storage module stores the pulse signal detection results according to storage related information, such as center frequency, bandwidth, and detection threshold. The high-speed data storage module uses an embedded solid-state disk array, and its architecture is as follows: Fig.13 shown.

[0129] Furthermore, part of the architecture of the entire broadband pulse signal acquisition and analysis system is designed based on the Direct Stream technology of the PCIe3.0 bus, such as Fig.14As shown, the broadband signal acquisition board refers to the hardware used in the broadband signal acquisition module, the real-time signal processing board refers to the signal processing and data analysis module, and the main controller refers to the human-computer interaction module. At the same time, it can be seen from the figure that the real-time signal processing board is set on the broadband signal acquisition board. The use of an efficient data flow controller can make data transmission more stable and efficient. It is built in RAID0 mode and high-speed data streaming is performed according to instrument control. A unique high-speed streaming control technology is used here. The measured results show that in 8x PCI Express Gen 3 mode, DMA Write (FPGA--> memory) can reach a speed of 6.2GB / s, while DMA Read (memory-->FPGA) performance is 6.0GB / s. The self-developed efficient data flow controller increases the efficiency of continuous data transmission by 5-10 percentage points, with obvious results. After a long period of optimization and testing and verification, the data storage system speed can reach more than 5GB / s.

[0130] In this embodiment, the human-computer interface between the broadband pulse signal acquisition system and the user is realized through the human-computer interaction module. The user can set various parameters through the graphical interface, control and manage the signal processing and data analysis module and the signal high-speed storage module, and display the results to the user in the form of a graphical interface.

[0131] In one embodiment, if Fig.15 As shown in the figure, a complete schematic diagram of the entire broadband pulse signal acquisition and analysis system is given. The broadband pulse signal acquisition and analysis system adopts a general PC platform and a 19-inch rack server. The overall hardware architecture is designed based on PCIe Gen3.0 and consists of a broadband downconverter (signal input and RF conversion module), FMC high-speed ADC acquisition daughter card (broadband signal acquisition module), PCIe high-performance FPGA real-time signal processing board and GPU processing card (signal processing and data analysis module), and a large-capacity solid-state storage array (signal high-speed storage module). The main technical indicators include: operating system: Windows 10, 64-bit, processor: Intel (R) Xeon processor, operating system hard disk: 500GB high-speed SSD hard disk, solid-state storage array: 8TB, memory: 16GB, GPU processing card: NVIDIA RTX3080, file system: NTFS, network interface: 1GbEthernet Port×2.

[0132] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0133] The above-mentioned embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the patent of the present application shall be subject to the attached claims.

Claims

1. A broadband pulse signal acquisition and analysis system, characterized in that: include: An input and radio frequency conversion module, a broadband signal acquisition module, a signal processing and data analysis module, a signal high-speed storage module connected in sequence, and a human-computer interaction module connected to the signal processing and data analysis module and the signal high-speed storage module respectively; The input and RF conversion module is used to receive broadband RF signals in real time, down-convert the RF broadband signals into intermediate frequency signals, and send the intermediate frequency signals to the broadband signal acquisition module; The broadband signal acquisition module is used to digitally sample the intermediate frequency signal, convert it into corresponding baseband signal IQ data, and send the baseband signal IQ data to the signal processing and data analysis module; The signal processing and data analysis module is used to perform real-time pulse detection and data afterglow statistics on the baseband signal IQ data according to data analysis related parameters to obtain pulse detection results, wherein when the baseband signal IQ data is subjected to real-time pulse detection, signal detection is performed on the baseband signal IQ data, and if a signal exists, time domain parameters and frequency domain parameters are detected; The signal high-speed storage module is used to store the detected pulse signal in real time according to the storage related information; The human-computer interaction module is used to obtain the data analysis related parameters and storage related information through the interactive interface, and send them to the signal processing and data analysis module and the signal high-speed storage module respectively, and also display the pulse detection results through the interactive interface.

2. The broadband pulse signal acquisition and analysis system according to claim 1, characterized in that: In the signal processing and data analysis module, when the baseband signal IQ data is subjected to real-time pulse detection and digital afterglow statistics to obtain pulse detection results according to data analysis related parameters: Perform signal detection on the baseband signal IQ data, and if a signal exists, perform time domain detection to obtain time domain parameters, or perform frequency domain detection using an instantaneous autocorrelation phase jump detection algorithm to obtain frequency domain parameters; Perform digital afterglow statistics according to the baseband signal IQ data to obtain digital afterglow statistics results; The time domain parameters, frequency domain parameters and digital afterglow statistics are used as the pulse detection results.

3. The broadband pulse signal acquisition and analysis system according to claim 2, characterized in that: When performing signal detection on the baseband signal IQ data: The baseband signal IQ data includes multiple frames of data sorted in time, and signal detection is performed on each frame of data in the baseband signal IQ data in turn; The GO constant false alarm rate detection method is used to perform pulse detection on the data after fast Fourier transformation. If a pulse signal is detected, a time domain or frequency domain detection is performed; If no pulse signal is detected, it is determined whether the number of frames with pulse edges before the current frame data exceeds half. If it exceeds half, the OS constant false alarm rate detection method is used to perform secondary pulse detection on the current frame data.

4. The broadband pulse signal acquisition and analysis system according to claim 3, characterized in that: When performing time domain detection on the baseband signal IQ data of the existing signal: After performing short-time Fourier transform on the baseband signal IQ data, the MN algorithm is used to perform coarse pulse edge detection. If a pulse edge is detected, a box differential detection filter is used to perform fine detection at the location of the pulse edge.

5. The broadband pulse signal acquisition and analysis system according to any one of claims 1 to 4, characterized in that: The frequency range of the real-time received broadband radio frequency signal is 1-18 GHZ, the center frequency of the intermediate frequency signal is 1.8 GHz, and the effective bandwidth is 1 GHz.

6. The broadband pulse signal acquisition and analysis system according to claim 5, characterized in that: The input and radio frequency conversion module includes multiple groups of pre-selection filter groups and a broadband down converter; The plurality of pre-selection filter groups filter the broadband radio frequency signals received in real time; The broadband down-converter uses its internally integrated low phase noise local oscillator to convert the filtered broadband RF signal into an intermediate frequency signal, and performs amplitude adjustment and anti-aliasing filtering on the intermediate frequency signal before sending it to the broadband signal acquisition module.

7. The broadband pulse signal acquisition and analysis system according to claim 5, characterized in that: The high-speed data storage module adopts an embedded solid-state disk array.

8. The broadband pulse signal acquisition and analysis system according to claim 5, characterized in that: The signal processing and data analysis module includes an FPGA unit and a GPU unit; In the FPGA unit, signal detection is performed on the baseband signal IQ data. If a signal exists, time domain detection is performed to obtain time domain parameters, and the pulse data is transmitted to the GPU unit through the PCle bus to perform digital afterglow statistics to obtain digital afterglow statistics results; In the GPU unit, frequency domain detection is performed using an instantaneous autocorrelation-based phase jump detection algorithm to obtain frequency domain parameters.

Citation Information

Patent Citations

  • Real-time pulse signal capture processing device and method

    CN104753563A

Cited By

  • Ultra-wideband software radio signal analysis device

    CN119519853A

  • Signal processing architecture of electro-optical front-end reconnaissance system

    CN119544083A

  • Real-time control and acquisition management system for pulse electromagnetic signal detection device

    CN119986549A