Pulse signal detection method and system
By converting the pulse signal into time domain energy data and processing it in segments, combining the detection methods of Z-fraction and adaptive sliding window step length, the problem of low pulse signal detection efficiency in the prior art is solved, and efficient and accurate pulse signal is achieved Screening and parameter estimation.
Patent Information
- Application Number
- CN202411121156.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-15
- Publication Date
- 2024-11-08
- Estimated Expiration
- 2044-08-15
AI Technical Summary
Existing pulse signal detection methods are inefficient when processing big data files and are difficult to accurately locate and estimate pulse signal parameters, especially in the absence of prior information.
By converting the signal to be detected into time domain energy data and performing smoothing processing, reading the data in segments, using the Z-fraction method to distinguish the pulse signal segment and the noise segment, and performing secondary detection based on the adaptive sliding window step length, improving detection efficiency and precision.
It significantly improves the efficiency and accuracy of pulse signal detection, reduces memory requirements, improves system stability and time resolution, reduces parameter estimation errors, and realizes rapid screening of pulse signal and parameter estimation.
Smart Images

Figure CN118916643A_ABST
Abstract
Description
Technical Field
[0001] The present disclosure relates to the technical field of signal analysis and identification, and in particular to a pulse signal detection method and system. Background Art
[0002] Pulse signal screening is widely used in complex electromagnetic signal analysis and identification. By screening the pulse signal of interest, the pulse signal of interest can be located and the parameters of the pulse signal can be estimated. However, there are many types of pulse signals, including linear frequency modulation (LFM), nonlinear frequency modulation (NLFM), phase-coded signal (PCS), etc. Pulse signal parameters are relatively large, including peak power, average power, start time, pulse width, pulse repetition interval, rise time, fall time, pulse frequency, and duty cycle. In practical applications, the in-phase and quadrature (IQ) data files formed by high-speed acquisition are often large and lack prior information, which brings great challenges to the accurate positioning and parameter estimation of pulse signals.
[0003] The research on pulse signal screening and parameter estimation technology is divided into two steps: signal detection and parameter estimation. Among them, in the field of signal detection, the Energy Detection Algorithm (EDA) is one of the most widely used methods in the field of wireless communication and signal processing. It can determine whether there is a target signal by calculating the time domain or frequency domain energy in the window and comparing it with the detection threshold. This method is simple and easy to implement and does not require prior knowledge, but it requires sliding processing of the energy window. The existing method uses the same calculation method for signals in all windows, and the screening efficiency is low when processing large data files. Summary of the invention
[0004] In order to solve the problems existing in the prior art, the embodiments of the present disclosure provide a pulse signal detection method and system to improve the efficiency of pulse signal detection. The technical solution is as follows: In a first aspect, a pulse signal detection method is provided, comprising: Acquire the signal to be detected, convert the signal to be detected into time domain energy data and perform smoothing processing; Divide the smoothed time-domain energy data into data segments of set length, perform preliminary pulse signal detection on each data segment, and divide the data segment into a pulse signal segment and a noise segment; The pulse signal segment is subjected to secondary detection based on the sliding window step size to obtain a detection result of the pulse signal segment.
[0005] In a second aspect, a pulse signal detection system is provided, comprising: A preprocessing module is configured to obtain a signal to be detected, convert the signal to be detected into time domain energy data and perform smoothing processing; A preliminary detection module is configured to divide the smoothed time-domain energy data into data segments of a set length, perform preliminary pulse signal detection on each data segment, and divide the data segment into a pulse signal segment and a noise segment; The secondary detection module is configured to perform secondary detection on the pulse signal segment based on the sliding window step size to obtain a detection result of the pulse signal segment.
[0006] In a third aspect, an electronic device is provided, comprising a memory, a processor, and a computer program stored in the memory, wherein the processor executes the computer program to complete the steps of the above-mentioned pulse signal detection method.
[0007] In a fourth aspect, a computer-readable storage medium is provided for storing computer instructions, which, when executed by a processor, complete the steps of the above-mentioned pulse signal detection method.
[0008] In a fifth aspect, a computer program product is provided, comprising a computer program / instruction, which, when executed by a processor, implements the steps of the above-mentioned pulse signal detection method.
[0009] The technical solution provided by the embodiment of the present disclosure has the following beneficial effects: in the embodiment of the present disclosure, data is first read in segments, and the pulse signal is regarded as a mutation value, and the Z score is used for detection to distinguish the pulse signal segment from the noise segment. In the subsequent secondary detection, only the data of the pulse signal segment is processed, thereby significantly improving the processing efficiency; by processing large data files in segments, it is avoided to load a large amount of data at one time, and the memory requirement is reduced, thereby improving the stability and efficiency of the system; for the pulse signal segment that may contain a pulse signal, an energy detection method based on an adaptive sliding window step size is adopted, and the sliding window step size is adaptively determined according to the characteristics of the pulse signal and the noise, which significantly improves the detection performance, and not only realizes the rapid screening of the pulse signal, but also reduces the error of the time parameter estimation by improving the time resolution.
[0010] Advantages of additional aspects of the present disclosure will be given in part in the following description and in part will become apparent from the following description or will be learned through practice of the present disclosure. BRIEF DESCRIPTION OF THE DRAWINGS
[0011] In order to more clearly illustrate the technical solutions in the embodiments of the present disclosure, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present disclosure. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.
[0012] Figure 1 is a flow chart of a pulse signal detection method provided by an embodiment of the present disclosure; Figure 2 The overall flow chart of the method is shown in FIG. Figure 3 Detailed flowchart of each step; Figure 4 Schematic diagram of fixed sliding window step size; Figure 5 Schematic diagram of adaptive sliding window step size; Figure 6 This is a schematic diagram comparing the number of processing times of the three methods. DETAILED DESCRIPTION
[0013] In order to make the objectives, technical solutions and advantages of the present disclosure more clear, the embodiments of the present disclosure will be further described in detail below with reference to the accompanying drawings.
[0014] The present disclosure provides a pulse signal detection method. Figure 1-Figure 3 , the method comprising: S101, obtaining a signal to be detected, converting the signal to be detected into time domain energy data and performing smoothing processing; S102, dividing the smoothed time-domain energy data into data segments of a set length, performing preliminary pulse signal detection on each data segment, and dividing the data segment into a pulse signal segment and a noise segment; S103 . Perform secondary detection on the pulse signal segment based on the sliding window step size to obtain a detection result of the pulse signal segment.
[0015] The disclosed embodiments accurately and quickly screen specific pulse signals contained in massive data, and in view of the characteristics of short pulse signal appearance time and long pulse signal-free period in complex electromagnetic environment, the traditional energy detection method adopts a fixed length of energy window sliding window step, and adopts the same processing method regardless of whether the signal intercepted by the window contains pulses, resulting in low screening efficiency. Pulse signal detection is performed based on mutation detection and adaptive sliding window step, and mutation value detection is performed on the segmented read signal to preliminarily determine whether the signal may contain pulses. For the signal segment that may contain pulses, the energy detection method based on adaptive sliding window step is used to detect the pulse signal, and its parameters are further estimated to achieve rapid screening and parameter estimation of pulse signals.
[0016] In step S101, for the IQ signal to be detected, the signal has the characteristics of amplitude, phase and frequency changing with time in the time domain. When performing signal detection and parameter estimation, converting the IQ signal into a square wave form has the advantages of simplifying calculation, facilitating estimation of pulse parameters, easy detection and energy concentration. In addition, the square wave signal has only two states, high and low levels, and the processing complexity is low. Therefore, the signal to be detected is converted into time domain energy data and smoothed. The time domain energy value is calculated using the following formula to convert the signal to be detected into time domain energy data: ; Among them, I and Q are the I-channel data and Q-channel data of the IQ data file respectively.
[0017] When converting the energy waveform into power form, the energy value is processed by a logarithmic function. The energy part less than 1 will decay rapidly and produce a sharp peak. Since the median filter can effectively retain the pulse edge and remove the sharp peak, the median filter is selected for smoothing. For the signal sample value at the center of the window , the output value of the median filter It can be obtained by the following formula: ; in, Represents the signal sample value The median value of .
[0018] After the IQ data is converted into energy data, the obtained data mainly consists of base noise and pulse signals in the form of square waves. In view of the fact that the period without pulse signals is usually long, this embodiment adopts a two-stage pulse signal detection method. First, it is determined whether the read data segment contains a pulse signal; then, the energy detection method based on the adaptive sliding window step size is applied to the data segment containing the pulse signal to detect the pulse signal, thereby significantly improving the signal processing efficiency.
[0019] Specifically, in step S102, the method for determining whether the read data segment contains a pulse signal is to read a data segment of length L each time and regard the signal therein as a mutation value. The data segment is analyzed by the Z score method to distinguish between the pulse signal segment and the noise segment.
[0020] The Z score is a standardized method for measuring the difference between a single data point and the mean of a data set. By calculating the mean and standard deviation of a data segment, the Z score can standardize whether a data point deviates from the mean, thereby effectively distinguishing between signal segments and noise segments. This method provides an accurate and reliable way to identify and deal with abnormal signals. , and its corresponding Z score is obtained by the following formula: ; in, is the average value of the data segment, is the standard deviation of the data segment.
[0021] Calculate the Z value of each point in turn and compare them. When the Z value of a point is greater than the set threshold, stop the calculation and determine that there is a pulse signal in the data segment. The data segment is a pulse signal segment; otherwise, the data segment is a noise segment. Record the number of read segments N of the pulse signal segment for subsequent time parameter estimation.
[0022] In step S103, the pulse signal segment is re-detected based on the adaptive sliding window step energy detection. The pulse signal segment determined by S102 contains not only the pulse signal, but also some noise. However, the duration of the pulse signal is relatively long, so the energy detection method based on the adaptive sliding window step is adopted here, and the sliding window step is dynamically adjusted, so that the signal detection process can be optimized according to different signal characteristics, thereby improving the accuracy and efficiency of detection.
[0023] First, according to the sampling frequency of the file and bandwidth Calculate the resolution bandwidth (RBW) and thus the window length for energy detection. Resolution bandwidth RBW and window length Calculated by the following formula: ; .
[0024] In order to improve the temporal resolution, the overlapping window method is usually used, and the initial sliding window step size is Set to The minimum energy value of the signal of interest is set as the detection threshold , the maximum value in the window, i.e. the peak power of the pulse signal With detection threshold The difference is used as the judgment basis. If there are data points with a difference greater than 0 in the window, it means that there is a pulse signal. If they are all less than 0, it means that the data in the window is base noise. By judging whether there is a pulse signal in the window, the sliding window step size of the next window is adaptively changed. The sliding window step size of the n+1th sliding is Use the sliding window step size of the nth sliding Multiplication adjustment factor Calculated, is an integer. The specific expression is: ; Since the initial sliding window step size , Unable to perform adaptive adjustment. The sliding window step length exceeds the window length, resulting in data loss, so here we take .
[0025] like Figure 4-Figure 5 By comparison, it can be seen that compared with the method of fixing the sliding window step size, this embodiment can adaptively determine the sliding window step size according to the characteristics of the pulse signal and noise, thereby significantly improving the detection performance.
[0026] In some implementations, the method further includes step S104 of performing parameter estimation on the detected pulse signal.
[0027] Since the signal usually appears continuously in multiple windows, when a complete pulse signal is detected, the single pulse is extracted for pulse parameter estimation. The above process is repeated until the entire IQ data file is processed, thereby generating a pulse signal parameter table, realizing rapid screening and parameter estimation of pulse signals in large data files.
[0028] Pulse parameters are divided into intra-pulse parameters and inter-pulse parameters. The parameter estimation of pulse signals includes intra-pulse parameter estimation and inter-pulse parameter estimation. Intra-pulse parameters include peak power , Start time , pulse width , Rise time , Fall time and pulse frequency The pulse-to-pulse parameters include the average power , pulse repetition interval and duty cycle Among them, the peak power It is the difference between the detection threshold and the energy detection value, which has been estimated during the signal detection process.
[0029] For the estimation of intrapulse parameters, the starting time The pulse width is the time point when the pulse changes from low level to high level, and the pulse width is the duration of the high level. Record each sliding window step size during adaptive sliding window step size and energy detection , and treat the signals appearing in multiple consecutive windows as a pulse and save them in the temporary pulse Recording signal start window number and end window number The start time of the signal and pulse width It can be obtained by the following formula: ; .
[0030] Rise time It is the time it takes for the pulse signal to rise from 10% to 90% of its maximum value, and the fall time It is the time it takes for the pulse signal to drop from 90% of its maximum value to 10%. and When the pulse signal is fully stored in First, we need to calculate The peak-to-peak value and the average power within the pulse width , thus determining the 10% threshold and 90% threshold .
[0031] Peak-to-Peak is the peak power and noise power of the pulse signal The difference between It is not an ideal square wave, but has fluctuations, so To characterize the maximum value of the pulse. and The calculation expression is as follows: ; .
[0032] and The intersection points of correspond to the end of the rise time and the beginning of the fall time, respectively. The intersection points correspond to the beginning of the rise time and the end of the fall time, respectively. It is a discrete digital signal. In order to save calculation time, it is not necessary to traverse all data points. Instead, a point-by-point comparison method is used from the peak power point to both sides. The peak value is traversed outwards. The first time the two sides are less than The points are the start of the rise time and the end of the fall time. From the start of the rise time and the end of the fall time, the first time it is less than The points are the end of the rise time and the beginning of the fall time. and The expression is as follows: ; .
[0033] in, , , , They are the rise time end index, the rise time start index, the fall time end index and the fall time end index.
[0034] The pulse frequency refers to the instantaneous frequency change of a single pulse signal on the time axis, which describes the frequency change characteristics of the pulse signal within a cycle. Due to the wide variety of pulse signals and large differences in spectrum, the center frequency of the pulse signal is used to represent its pulse frequency. .
[0035] For the estimation of pulse-to-pulse parameters, the pulse-to-pulse parameters have the average power , pulse repetition interval and duty cycle The core of the pulse-to-pulse parameter is the pulse repetition interval , as long as we calculate By combining the calculated intra-pulse parameters with the values of , other inter-pulse parameters can be derived.
[0036] To calculate the pulse repetition interval, you need to add a variable saver , q is the number of pulse signals. The calculation formula is as follows: .
[0037] Duty Cycle Can be based on and calculate, Can be based on and Calculation, the calculation formula is as follows: ; .
[0038] The effectiveness of the method provided by the embodiment of the present disclosure is illustrated below by taking a specific signal as an example.
[0039] LFM signal is one of the most common signal types in pulse signal processing. In order to verify the effectiveness of the disclosed method, five linear frequency modulation (LFM) pulse signals with different parameters were selected for simulation experiments, and the pulse signal data generated by the simulation was compared with the results of the method. The sampling frequency is 2.5e9, the number of data points is 5e6, the resolution bandwidth (RBW) is 1MHz, and the pulse parameters in the 100~200us time period are: =20us, =4us, =0.2, =0.4GHz, period is 5; the pulse parameters in the 230~260us time period are: =10us, =3.5us, =0.1, =0.2GHz, period is 3; the pulse parameters in the 336~396us time period are: =12us, =8us, =2 / 3, =0.3GHz, period is 5; the pulse parameters in the time period of 1245~1365us are: =30us, =7.5us, =0.25, =0.15GHz, period is 4; the pulse parameters in the 1396~1546us time period are: =50us, =5.5us, =0.11, =1GHz, period is 3.
[0040] The window length is set by the sampling frequency and the resolution bandwidth. Under the premise of keeping the window length unchanged, the number of times processed by the following three energy detection methods is compared: the classic energy detection method, the energy detection method based on the adaptive sliding window step, and the energy detection method based on the adaptive sliding window step after first reading the data in segments to determine the pulse signal segment. The comparison results are as follows: Figure 6 shown.
[0041] Figure 6 The horizontal axis in the middle represents the proportion of pulse signals in the IQ data file, while the vertical axis shows the total number of processing times, that is, the number of energy windows. In the classic energy detection method, the window length and step length are both fixed values, and the number of processing times of the simulation test file is fixed at 3999 times. For the energy detection method based on adaptive sliding window step length, the number of processed windows is negatively correlated with the proportion of the signal: that is, the higher the signal proportion, the more windows are processed; the lower the proportion, the fewer windows. In the energy detection method based on adaptive sliding window step length, which first determines the existence of the signal and then applies the energy detection method based on adaptive sliding window step length, the number of windows is also negatively correlated with the signal proportion. In addition, the method of first determining the signal does not perform energy detection on the noise segment without signal, so that the number of windows is minimized, which significantly improves the processing efficiency and speeds up the screening of pulse signals.
[0042] After screening out the pulse signal, the next step is parameter estimation. The estimation results are shown in Table 1.
[0043] Table 1 Pulse signal screening and parameter estimation results
[0044] Analyzing the data in the table, the average error of the start time is 0.23us; the average error of the pulse width is 0.001us; the average error of the repetition frequency is 0.1875us (the repetition frequency is the difference between the start times of adjacent pulses. The alternation of pulse sequences with different periods will cause a significant change in the repetition interval of the last pulse of the previous sequence, which does not participate in the error calculation); the average error of the pulse frequency is 2.05kHz. For the time-frequency resolution of high-frequency transient signals, the error of the resulting parameters is small. It can be seen that the method provided by the embodiment of the present disclosure not only improves the processing efficiency, but also ensures the high accuracy of parameter estimation.
[0045] The present disclosure further provides a pulse signal detection system, the system comprising: A preprocessing module is configured to obtain a signal to be detected, convert the signal to be detected into time domain energy data and perform smoothing processing; A preliminary detection module is configured to divide the smoothed time-domain energy data into data segments of a set length, perform preliminary pulse signal detection on each data segment, and divide the data segment into a pulse signal segment and a noise segment; The secondary detection module is configured to perform secondary detection on the pulse signal segment based on the sliding window step size to obtain a detection result of the pulse signal segment.
[0046] The pulse signal detection system provided in the above embodiment is illustrated by the division of the above hardware modules when performing pulse signal detection. In practical applications, the above hardware modules can be replaced with other hardware modules with similar or identical functions as needed, or the functions of a certain hardware module can be assigned to different hardware modules. In addition, the above embodiment is the same as the implementation method of a pulse signal detection method, which will not be repeated here.
[0047] The embodiment of the present disclosure further provides an electronic device, including a memory, a processor, and a computer program stored in the memory, wherein the processor executes the computer program to complete the steps of the method provided above.
[0048] The processor may include one or more processing cores, such as a 4-core processor, an 8-core processor, etc. The processor may be implemented in at least one hardware form of DSP (Digital Signal Processing), FPGA (Field-Programmable Gate Array), and LA (Programmable Logic Array). The processor may also include a main processor and a coprocessor. The main processor is a processor for processing data in the awake state, also known as a CPU (Central Processing Unit); the coprocessor is a low-power processor for processing data in the standby state. In some embodiments, the processor may be integrated with a GPU (Graphics Processing Unit), which is responsible for rendering and drawing the content to be displayed on the display screen. In some embodiments, the processor may also include an AI (Artificial Intelligence) processor, which is used to process computing operations related to machine learning.
[0049] The memory may include one or more computer-readable media, which may be non-transitory. The memory may also include high-speed random access memory, and non-volatile memory, such as one or more disk storage devices, flash memory storage devices. In some embodiments, the non-transitory computer-readable medium in the memory is used to store at least one computer program, which is used to be executed by the processor to implement a pulse signal detection method provided in an embodiment of the present disclosure.
[0050] The embodiment of the present disclosure also provides a computer-readable storage medium for storing computer instructions. When the computer instructions are executed by a processor, the steps of a pulse signal detection method provided in the embodiment of the present disclosure can be completed.
[0051] The embodiment of the present disclosure also provides a computer program product, including a computer program / instruction, characterized in that when the computer program / instruction is executed by a processor, the steps of a pulse signal detection method provided in the embodiment of the present disclosure are implemented.
[0052] The above description is only a preferred embodiment of the present disclosure and is not intended to limit the present disclosure. For those skilled in the art, the present disclosure may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present disclosure shall be included in the protection scope of the present disclosure.
Claims
1. A pulse signal detection method, characterized in that: include: Acquire the signal to be detected, convert the signal to be detected into time domain energy data and perform smoothing processing; Divide the smoothed time-domain energy data into data segments of set length, perform preliminary pulse signal detection on each data segment, and divide the data segment into a pulse signal segment and a noise segment; The pulse signal segment is subjected to secondary detection based on the sliding window step size to obtain a detection result of the pulse signal segment.
2. A pulse signal detection method as claimed in claim 1, characterized in that: The smoothing process includes performing median filtering on the time domain energy data.
3. A pulse signal detection method as claimed in claim 1, characterized in that: The method of performing preliminary pulse signal detection on each data segment and dividing the data segment into a pulse signal segment and a noise segment includes: based on the Z score, calculating the Z value of each point in each data segment in turn, if the Z value of any point is greater than a set threshold, dividing the data segment where the point is located into a pulse signal segment; if the Z value of each point in the data segment is not greater than the set threshold, dividing the data segment into a noise segment.
4. A pulse signal detection method as claimed in claim 1, characterized in that: The secondary detection of the pulse signal segment based on the sliding window step size includes: calculating the resolution bandwidth and window length according to the sampling frequency and bandwidth; setting the initial sliding window step size based on the window length; determining the sliding window step size of the next window according to the comparison result of the peak power of the pulse signal in the previous window and the detection threshold; determining the next window according to the sliding window step size of the next window, and detecting whether the data in each window is a pulse signal.
5. A pulse signal detection method as claimed in claim 4, characterized in that: The sliding window step size of the next window is calculated based on the comparison result of the peak power of the pulse signal in the previous window and the detection threshold, the sliding window step size of the previous window and the adjustment factor.
6. A pulse signal detection method as claimed in claim 1, characterized in that: The method also includes performing parameter estimation on the detected pulse signal, including intra-pulse parameter estimation and inter-pulse parameter estimation.
7. A pulse signal detection system, characterized in that: include: A preprocessing module is configured to obtain a signal to be detected, convert the signal to be detected into time domain energy data and perform smoothing processing; A preliminary detection module is configured to divide the smoothed time-domain energy data into data segments of a set length, perform preliminary pulse signal detection on each data segment, and divide the data segment into a pulse signal segment and a noise segment; The secondary detection module is configured to perform secondary detection on the pulse signal segment based on the sliding window step size to obtain a detection result of the pulse signal segment.
8. An electronic device, characterized in that: The invention comprises a memory, a processor and a computer program stored in the memory, wherein the processor executes the computer program to complete the steps of the method according to any one of claims 1 to 6.
9. A computer-readable storage medium, characterized in that: Used to store computer instructions, which, when executed by a processor, complete the steps of the method according to any one of claims 1 to 6.
10. A computer program product comprising a computer program / instructions, characterized in that When the computer program / instructions are executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.
Citation Information
Patent Citations
Fast identification device for continuous wave and pulse signals
CN109167746A
Radar respiration signal detection method and system based on walking state
CN114642417A
Impulse noise suppression method and system, readable storage medium and computer equipment
CN115348507A
Impulse noise suppression method in NOMA communication system
CN116723073A
System and method for detecting pulses using fused signal power / phase modulation detection
US10725151B1
Cited By
Pulsed arc spectrum component measuring device and method based on adjustable sliding window
CN119595107A
Pulse arc spectrum component measurement device and method based on adjustable sliding window
CN119595107B
