A distributed radar target detection method and system based on single-bit data

Abstract

The present invention discloses a distributed radar target detection method and system based on single-bit data, the method comprising: obtaining a target received signal, and performing single-bit quantization processing on the target received signal to obtain a single-bit digital signal; performing pulse division processing and matched filtering processing on the single-bit digital signal to obtain a pulse compression signal; obtaining a column of data with the largest power and a time delay estimation value in the pulse compression signal, and performing Doppler frequency estimation processing based on the column of data with the largest power to obtain multiple Doppler frequency estimation values; performing parameter compensation processing based on the time delay estimation value and the multiple Doppler frequency estimation values ​​to obtain a cumulative test statistic, and when the cumulative test statistic is greater than a preset detection threshold, it is determined that the target detection object exists. The present invention performs parameter compensation through time delay and Doppler frequency, which can achieve a more reliable target detection task.

Description

Technical Field

[0001] The present invention relates to the field of radar signal processing technology, and in particular to a distributed radar target detection method, system, terminal and computer-readable storage medium based on single-bit data. Background Art

[0002] Distributed radar systems achieve wider surveillance coverage and multi-angle target detection by breaking down traditional centralized large radars into multiple smaller sites and distributing them geographically, demonstrating significant value in the field of target detection. However, distributed radar systems also bring new challenges, including high hardware implementation complexity and large data transmission bandwidth consumption.

[0003] To solve these problems, single-bit sampling technology has emerged as a promising solution. This technology uses a single-bit quantizer to convert analog signals into binary form of +1 or -1, which greatly reduces hardware complexity and cost, reduces energy consumption, and increases the sampling rate. These advantages make single-bit radar very suitable for distributed radar systems, especially on small platforms such as drones.

[0004] However, the detection process of the target object by using the single-bit sampling technology in the prior art does not take into account the time delay and Doppler information in actual applications, resulting in inaccurate detection results of the existence of the target object.

[0005] Therefore, the prior art still needs to be improved and developed. Summary of the invention

[0006] The main purpose of the present invention is to provide a distributed radar target detection method, system, terminal and computer-readable storage medium based on single-bit data, aiming to solve the problem that the detection process of the target object by single-bit sampling technology in the prior art does not take into account the delay and Doppler information in actual application, thereby resulting in inaccurate detection results of the target object's existence.

[0007] To achieve the above object, the present invention provides a distributed radar target detection method based on single-bit data, and the distributed radar target detection method based on single-bit data comprises the following steps:

[0008] Acquire a target received signal, and perform single-bit quantization processing on the target received signal to obtain a single-bit digital signal;

[0009] The single-bit digital signal is subjected to pulse division processing and matched filtering processing to obtain a pulse compression signal;

[0010] Acquire a column of data with the largest power and a time delay estimation value in the pulse compression signal, and perform Doppler frequency estimation processing according to the column of data with the largest power to obtain multiple Doppler frequency estimation values;

[0011] Parameter compensation processing is performed according to the time delay estimation value and the multiple Doppler frequency estimation values ​​to obtain a cumulative test statistic. When the cumulative test statistic is greater than a preset detection threshold, it is determined that the target detection object exists.

[0012] Optionally, the distributed radar target detection method based on single-bit data, wherein the step of acquiring a target received signal and performing single-bit quantization processing on the target received signal to obtain a single-bit digital signal, specifically includes:

[0013] A pulse signal is transmitted through a radar transmitter, and a returned target receiving signal is received through a radar receiver, wherein the expression of the target receiving signal is:

[0014] ;

[0015] in, For the r The target receiving signal received by the radar receiver is for No. m The target received signal corresponding to each sampling time point is For the r The complex reflection coefficient corresponding to the target reception signal received by a radar receiver is For the quantifier, is the number of pulse signals transmitted, is a single pulse signal, For the r The delay corresponding to the target receiving signal is is the sampling length, is the imaginary unit, For the r The Doppler frequency corresponding to the target reception signal received by a radar receiver, is complex Gaussian white noise, For the m Complex Gaussian white noise corresponding to the sampling time point;

[0016] The target received signal is subjected to single-bit quantization processing to obtain a single-bit digital signal, wherein the expression of the single-bit digital signal is:

[0017] ;

[0018] in, For the rThe target receiving signal received by a radar receiver is a single-bit digital signal after single-bit quantization. is the symbolic function, To obtain the real part, To take the imaginary part.

[0019] Optionally, the distributed radar target detection method based on single-bit data, wherein the single-bit digital signal is subjected to pulse division processing and matched filtering processing to obtain a pulse compression signal, specifically includes:

[0020] Performing pulse division processing on the single-bit digital signal to obtain a received pulse signal;

[0021] A transmission pulse signal is obtained, and a matched filtering algorithm is used to perform matched filtering on the transmission pulse signal and the received pulse signal to obtain a pulse compression signal, wherein the expression of the pulse compression signal is:

[0022] ;

[0023] in, is the pulse compression signal, The first bit of the single-bit digital signal after pulse division processing A pulse signal, is the convolution calculation, is the conjugate transpose.

[0024] Optionally, the distributed radar target detection method based on single-bit data, wherein the step of acquiring a column of data with the largest power in the pulse compression signal and a delay estimation value, and performing Doppler frequency estimation processing according to the column of data with the largest power to obtain multiple Doppler frequency estimation values, specifically includes:

[0025] Find a column of data with the largest power and the corresponding maximum power position in the pulse compression signal, and use the maximum power position as the delay estimation value, wherein the expression of the maximum power position is:

[0026] ;

[0027] in, is the maximum power position, Solving for the maximum power position, , , is the power data of each column in the pulse compression signal, For the column vector, is the square of the vector's two-norm;

[0028] Perform Fourier transform processing on the column of data with the largest power, perform spectrum peak search in the frequency domain, obtain a first spectrum peak position, and obtain a first Doppler frequency according to the first spectrum peak position, wherein the expression of the first Doppler frequency is:

[0029] ;

[0030] in, is the first Doppler frequency, is the position of the first spectral peak;

[0031] Performing fine frequency compensation processing on the first spectrum peak position to obtain a second Doppler frequency;

[0032] A sample covariance matrix is ​​constructed according to the target received signal and the time delay estimation value, and a multiple signal classification algorithm is used to perform Doppler frequency estimation processing on the sample covariance matrix to obtain a third Doppler frequency.

[0033] Optionally, the distributed radar target detection method based on single-bit data, wherein the performing of frequency fine compensation processing on the first spectrum peak position to obtain the second Doppler frequency specifically includes:

[0034] The first spectrum peak position is subjected to fine frequency compensation processing by using a fine search Fourier transform algorithm to obtain a second spectrum peak position, wherein the expression of the second spectrum peak position is:

[0035] ;

[0036] in, is the second peak position, Solve for the position of the second peak, is the corresponding rotation phase, is the standard Fourier transform matrix OK, is the phase rotation of the original vector, is the column vector corresponding to the maximum power index;

[0037] A second Doppler frequency is obtained according to the second spectrum peak position, wherein the expression of the second Doppler frequency is:

[0038] ;

[0039] in, is the second Doppler frequency.

[0040] Optionally, the distributed radar target detection method based on single-bit data, wherein the step of constructing a sample covariance matrix according to the target received signal and the delay estimation value, and performing Doppler frequency estimation processing on the sample covariance matrix using a multiple signal classification algorithm to obtain a third Doppler frequency, specifically includes:

[0041] Constructing a sample covariance matrix according to the target received signal and the delay estimation value, and performing eigenvalue decomposition processing on the sample covariance matrix to obtain a noise subspace;

[0042] A spectrum search is performed according to the noise subspace to obtain a third spectrum peak position, and a third Doppler frequency is obtained according to the third spectrum peak position.

[0043] Optionally, the distributed radar target detection method based on single-bit data, wherein the parameter compensation processing is performed according to the time delay estimation value and the multiple Doppler frequency estimation values ​​to obtain a cumulative test statistic, and when the cumulative test statistic is greater than a preset detection threshold, it is determined that the target detection object exists, specifically including:

[0044] The delay estimation value, the first Doppler frequency, the second Doppler frequency, and the third Doppler frequency are input into matched filter detectors of multiple radar receivers, and multiple initial test statistics are output, wherein the expression of the initial test statistics is:

[0045] ;

[0046] in, For the r The initial test statistic of a radar receiver, tr is the trace of the matrix, For the r The complex matrix signal corresponding to a radar receiver is , For the r The delay estimate corresponding to each radar receiver is: For the r a first Doppler frequency, a second Doppler frequency, and a third Doppler frequency corresponding to a radar receiver;

[0047] Integrating the multiple initial test statistics output by the multiple radar receivers using a non-coherent accumulation method to obtain a cumulative test statistic;

[0048] The accumulated test statistic is compared with a preset detection threshold, and when the accumulated test statistic is greater than the preset detection threshold, it is determined that the target detection object exists.

[0049] In addition, to achieve the above-mentioned purpose, the present invention further provides a distributed radar target detection system based on single-bit data, wherein the distributed radar target detection system based on single-bit data includes:

[0050] A single-bit quantization processing module is used to obtain a target received signal and perform single-bit quantization processing on the target received signal to obtain a single-bit digital signal;

[0051] A pulse compression signal generating module, used for performing pulse division processing and matched filtering processing on the single-bit digital signal to obtain a pulse compression signal;

[0052] A time delay and Doppler frequency acquisition module, used to acquire a column of data with the largest power and a time delay estimation value in the pulse compression signal, and perform Doppler frequency estimation processing based on the column of data with the largest power to obtain multiple Doppler frequency estimation values;

[0053] The target detection object judgment module is used to perform parameter compensation processing according to the delay estimation value and the multiple Doppler frequency estimation values ​​to obtain a cumulative test statistic. When the cumulative test statistic is greater than a preset detection threshold, it is determined that the target detection object exists.

[0054] In addition, to achieve the above-mentioned purpose, the present invention also provides a terminal, wherein the terminal includes: a memory, a processor, and a distributed radar target detection program based on single-bit data stored in the memory and executable on the processor, wherein the distributed radar target detection program based on single-bit data implements the steps of the distributed radar target detection method based on single-bit data as described above when executed by the processor.

[0055] In addition, to achieve the above-mentioned purpose, the present invention also provides a computer-readable storage medium, wherein the computer-readable storage medium stores a distributed radar target detection program based on single-bit data, and when the distributed radar target detection program based on single-bit data is executed by a processor, the steps of the distributed radar target detection method based on single-bit data as described above are implemented.

[0056] In the present invention, a target received signal is obtained, and the target received signal is subjected to single-bit quantization processing to obtain a single-bit digital signal; the single-bit digital signal is subjected to pulse division processing and matched filtering processing to obtain a pulse compression signal; a column of data with the largest power and a time delay estimate in the pulse compression signal are obtained, and a Doppler frequency estimate is performed according to the column of data with the largest power to obtain multiple Doppler frequency estimates; parameter compensation processing is performed according to the time delay estimate and multiple Doppler frequency estimates to obtain a cumulative test statistic, and when the cumulative test statistic is greater than a preset detection threshold, it is determined that the target detection object exists. The present invention not only reduces the burden of distributed radar data transmission, but also solves the problem of accuracy of parameter recovery of single-bit technology under low signal-to-noise ratio conditions by subjecting the single-bit digital signal obtained after single-bit quantization processing to pulse division processing and matched filtering processing, thereby determining the maximum power data and the maximum power position, and further calculating the time delay and Doppler frequency. At the same time, through the parameter compensation of time delay and Doppler frequency, a more reliable target detection task is achieved. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] Figure 1 It is a flow chart of a preferred embodiment of the distributed radar target detection method based on single-bit data of the present invention;

[0058] Figure 2 Schematic diagram of the relationship between the mean absolute error of delay estimation and the signal-to-noise ratio of a preferred embodiment of the distributed radar target detection method based on single-bit data of the present invention;

[0059] Figure 3 It is a schematic diagram of the relationship between the mean square error of Doppler frequency estimation and the signal-to-noise ratio of a preferred embodiment of the distributed radar target detection method based on single-bit data of the present invention;

[0060] Figure 4 Schematic diagram of radar receiver operating characteristic curve of a preferred embodiment of the distributed radar target detection method based on single-bit data of the present invention;

[0061] Figure 5 It is a structural diagram of a preferred embodiment of the distributed radar target detection system based on single-bit data of the present invention;

[0062] Figure 6 It is a structural diagram of a preferred embodiment of the terminal of the present invention. DETAILED DESCRIPTION

[0063] In order to make the purpose, technical solution and advantages of the present invention clearer and more specific, the present invention is further described in detail below with reference to the accompanying drawings and examples. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0064] Distributed radar systems achieve wider surveillance coverage and multi-angle target detection by breaking down traditional centralized large radars into multiple smaller sites and distributing them geographically, demonstrating significant value in the field of target detection. However, distributed radar systems also bring new challenges, including high hardware implementation complexity and transmission bandwidth consumption of large amounts of data. To address these issues, one-bit sampling technology (also called single-bit sampling technology) has emerged as a promising solution. This technology uses a one-bit quantizer to convert analog signals into binary form of +1 or -1, which greatly reduces hardware complexity and cost, reduces energy consumption, and increases the sampling rate. These advantages make one-bit radar very suitable for distributed radar systems, especially on small platforms such as drones. In the past two decades, studies have shown that one-bit sampling technology has been widely verified in radar applications, including scattering coefficient estimation, arrival direction estimation, frequency estimation, and distance estimation. These studies highlight the importance of parameter estimation in the effective application of one-bit radar. In the field of target detection, it is shown in the prior art that one-bit data has a performance loss of about 2dB compared to multi-bit data at the same sample size. However, there is unknown delay and Doppler information in practical applications. However, studies have shown that ignoring this information will lead to a decrease in detector performance due to mismatch. Therefore, in the context of one-bit radar, it is necessary to select an appropriate signal processing algorithm for parameter estimation to enhance detection performance. The present invention aims to overcome the limitations of the prior art by studying a single-bit parameter estimation method, achieve robust multi-parameter joint estimation, provide accurate support for target detection, improve detector performance through parameter compensation, and reveal the positive correlation between parameter estimation accuracy and target detection effect.

[0065] To solve the above problems, the present invention proposes a distributed radar target detection method based on single-bit data, which reduces the burden of distributed radar data transmission and solves the accuracy problem of parameter recovery of single-bit technology under low signal-to-noise ratio conditions. At the same time, through parameter compensation of time delay and Doppler frequency, a more reliable target detection task is achieved.

[0066] The distributed radar target detection method based on single-bit data described in the preferred embodiment of the present invention is as follows: Figure 1 As shown, the distributed radar target detection method based on single-bit data includes the following steps:

[0067] Step S10: Acquire a target received signal, and perform single-bit quantization processing on the target received signal to obtain a single-bit digital signal.

[0068] The present invention firstly collects the analog baseband signal through a single-bit analog-to-digital converter (ADC) to obtain a single-bit digital signal.

[0069] Specifically, a pulse signal is transmitted through a radar transmitter, and a returned target reception signal is received through a radar receiver, wherein the expression of the target reception signal is: ;in, For the r The target receiving signal received by the radar receiver is for No. m The target received signal corresponding to each sampling time point is For the r The complex reflection coefficient corresponding to the target reception signal received by a radar receiver is For the quantifier, is the number of pulse signals transmitted, is a single pulse signal, For the r The delay corresponding to the target receiving signal is is the sampling length, is the imaginary unit, For the r The Doppler frequency corresponding to the target reception signal received by a radar receiver, is complex Gaussian white noise, For the m The complex Gaussian white noise corresponding to the sampling time point; the target received signal is subjected to single-bit quantization processing to obtain a single-bit digital signal, wherein the expression of the single-bit digital signal is:

[0070] ;in, For the r The target receiving signal received by a radar receiver is a single-bit digital signal after single-bit quantization. is the symbolic function, To obtain the real part, To take the imaginary part.

[0071] It is understandable that, for target detection, the present invention considers the case of using 1 radar transmitter, 1 moving target, 2 radar receivers, and known transmission signals. Since the distributed radar receivers are placed at different distances and angles, the corresponding time delays, Doppler frequencies, and reflection coefficients are all different.

[0072] The radar transmitter sends pulse signals to the moving target, continuously Pulse signal, The sampling time points are expressed as: ;in, , is a complex matrix, , Indicates that a sampling length is The single pulse signal, The sampling time points can be expressed as ,in is the length of the useful signal sampled, is the transmitting frequency.

[0073] The signal reflected by the target passes through the r After the receiving unit of the radar receiver samples, the m The target received signal corresponding to the sampling time point can be expressed as ;in, For the r The delay corresponding to the target receiving signal is ,make , It is r The target receiving signal received by the radar receiver is is complex Gaussian white noise, For the r The complex reflection coefficient corresponding to the target reception signal received by a radar receiver is For the r The Doppler frequency corresponding to the target received signal received by a radar receiver. Further, the received echo signal (i.e., the target received signal in the present invention) is single-bit quantized, and the received signal after single-bit quantization (i.e., the single-bit digital signal in the present invention) is: ,in, It is r The target receiving signal received by a radar receiver is a single-bit digital signal after single-bit quantization. is a sign function, is the imaginary unit, is to take the real part, It is to take the imaginary part.

[0074] The task of the present invention is to identify whether a target exists based on quantitative observations. is the assumption that there is a target in the received data, and Assume that there is no target. Without quantization, the standard target detection problem can be described as a binary hypothesis test: ,in is the target received signal in the noise-free case, , and when using single-bit technology, the target detection problem becomes: .

[0075] Step S20: Perform pulse division processing and matched filtering processing on the single-bit digital signal to obtain a pulse compression signal.

[0076] The present invention divides the signal after a single bit (ie, the single-bit digital signal in the present invention) into pulses, and adopts a matched filtering method for each receiving pulse and transmitting pulse to realize pulse compression.

[0077] Specifically, the single-bit digital signal is pulse divided to obtain a received pulse signal; a transmitted pulse signal is obtained, and a matched filtering algorithm is used to perform matched filtering on the transmitted pulse signal and the received pulse signal to obtain a pulse compression signal, wherein the expression of the pulse compression signal is: ;in, is the pulse compression signal, The first bit of the single-bit digital signal after pulse division processing A pulse signal, , is the convolution calculation, is the conjugate transpose.

[0078] The present invention converts the received signal after single-bit quantization processing (i.e., the single-bit digital signal in the present invention) According to different pulse numbers, Dimensional matrix , The row vector of , pulse compression is achieved by using matched filtering for each received pulse and transmitted pulse, which can be expressed as: , where the symbol represents the convolution calculation, represents the conjugate transpose, , is the output after piecewise matched filtering.

[0079] Step S30, obtaining a column of data with the largest power and a time delay estimation value in the pulse compression signal, and performing Doppler frequency estimation processing according to the column of data with the largest power to obtain multiple Doppler frequency estimation values.

[0080] The present invention estimates the time delay by finding the position of a column of data with the largest power after pulse compression, and then further selects a column of data with the largest power after pulse compression for FFT to obtain the position of the spectrum peak, thereby roughly estimating the Doppler frequency. Afterwards, a fine search is performed around the spectrum peak through fine frequency compensation, and the frequency corresponding to the new spectrum peak is converted into an estimated value of the Doppler frequency. Finally, a sample covariance matrix is ​​constructed, and the Doppler frequency estimation is completed through the MUSIC algorithm.

[0081] Specifically, a column of data with the largest power and a corresponding maximum power position are found in the pulse compression signal, and the maximum power position is used as the delay estimation value, wherein the expression of the maximum power position is:

[0082] ;in, is the maximum power position, Solve for the maximum power position (i.e. find the parameter To find the maximum value of the function, Max is to find a ,this Make This function has a maximum value), , , is the power data of each column in the pulse compression signal, For the column vector, is the square of the vector two norm; performing Fourier transform processing on the column of data with the largest power, performing spectrum peak search in the frequency domain, obtaining a first spectrum peak position, and obtaining a first Doppler frequency according to the first spectrum peak position, wherein the expression of the first Doppler frequency is:

[0083] ;in, is the first Doppler frequency, is the position of the first spectral peak.

[0084] The output signal after pulse compression reflects the characteristics of the target, and the maximum power point corresponds to the strongest target signal. Therefore, it is the best choice to select the output signal at the maximum power position for parameter estimation. yes The matrix composed of the results of segmented pulse matching filtering, The column vector of can be expressed as , assuming is the index of the maximum power in the corresponding column vector (i.e., the maximum power position in the present invention). Obviously, the position with the maximum power is the optimal delay estimate. Then we have ; ;in is the maximum power position, is the estimated delay (i.e., the delay corresponding to the rth target received signal in the present invention), Represents the square of the two-norm of a vector.

[0085] Furthermore, the column of data with the largest power after pulse compression is selected for FFT (Fourier transform). is the standard Fourier transform matrix, No. The elements can be represented as ,in, ,and , assuming is the column vector corresponding to the maximum power index, expressed as , , Middle The elements can be represented as, , , It is r The Doppler frequency of the target signal received by a radar receiver corresponds to It follows The change can be expressed as , so the Doppler frequency can be estimated by searching for the spectrum peak position in the frequency domain (i.e., the local maximum of FFT), and the position of the spectrum peak can be expressed as: ; The initial value of the Doppler frequency can be roughly estimated : , thus completing a rough estimate of the Doppler frequency.

[0086] The first spectrum peak position is subjected to fine frequency compensation processing by using a fine search Fourier transform algorithm to obtain a second spectrum peak position, wherein the expression of the second spectrum peak position is: ;in, is the second peak position, Solve for the second peak position (i.e. find the parameter To find the maximum value of the function, argmax is to find a parameter value, so that This function reaches its maximum value). is the corresponding rotation phase, is the standard Fourier transform matrix OK, is the phase rotation of the original vector, is the column vector corresponding to the maximum power index; and the second Doppler frequency is obtained according to the second spectrum peak position, wherein the expression of the second Doppler frequency is: ;in, is the second Doppler frequency.

[0087] The present invention uses a fine search FFT algorithm to achieve accurate estimation of Doppler frequency. The size of the frequency represented by a grid on the FFT frequency axis is expressed as , then the frequency points can be refined near the first spectrum peak position, and then the peak position can be fine-tuned to maximize the peak, thus overcoming the resolution limitation of FFT. The phase rotation of the original vector is defined as , ,in, is the corresponding rotation phase, range , the position of the new maximum peak can be expressed as: ;in yes No. OK, Therefore, after obtaining a new peak maximum position, a more accurate Doppler frequency estimation can be achieved. : .

[0088] A sample covariance matrix is ​​constructed according to the target received signal and the time delay estimation value, and an eigenvalue decomposition process is performed on the sample covariance matrix to obtain a noise subspace; a spectrum search is performed according to the noise subspace to obtain a third spectrum peak position, and a third Doppler frequency is obtained according to the third spectrum peak position.

[0089] Construct the sample covariance matrix and estimate the Doppler frequency using the MUSIC algorithm. Cut according to different pulse numbers dimensional matrix, using the estimated time delay, extract the useful signal after the estimated time delay to form Dimensional receive array matrix , which can be expressed as: , The relationship between the elements in can be expressed as , obviously It reflects the phase change of the target received signal at different Doppler frequencies and can be written as Therefore, the MUSIC algorithm can be used to estimate the Doppler frequency, and the matrix after receiving a single bit of the signal can be written as: ;in, It means that the received signal after single-bit quantization is cut into different pulse numbers dimensional matrix, using the estimated time delay, extract the useful signal after the estimated time delay to form dimensional receive array matrix, so the covariance matrix of the receive data It is expressed as: .

[0090] The following are the general steps to estimate Doppler frequency using the MUSIC algorithm:

[0091] 1. Estimate the one-bit covariance matrix as: ; 2. Perform feature decomposition: , is the covariance matrix The eigenvector matrix of It can be divided into two parts: the eigenvector of the signal subspace and the eigenvector of the noise subspace; 3. Output the noise subspace matrix , that is, 4. Perform spectrum search: ; 5. Output Doppler estimation based on the position of the highest peak in the spectrum .

[0092] It should be noted that the first spectrum peak, the second spectrum peak and the third spectrum peak in the present invention are not named according to the size of the spectrum peak, but are only used to distinguish the spectrum peaks calculated by different algorithms. Similarly, the first Doppler frequency, the second Doppler frequency and the third Doppler frequency are also the same. That is, the first spectrum peak and the first Doppler frequency correspond to the FFT (Fourier) algorithm; the second spectrum peak and the second Doppler frequency correspond to the Fourier transform algorithm of fine search; the third spectrum peak and the third Doppler frequency correspond to the multiple signal classification algorithm. The present invention obtains the first Doppler frequency, the second Doppler frequency and the third Doppler frequency through different algorithms and compares their respective accuracies.

[0093] Step S40: Perform parameter compensation processing according to the time delay estimation value and the multiple Doppler frequency estimation values ​​to obtain a cumulative test statistic. When the cumulative test statistic is greater than a preset detection threshold, it is determined that the target detection object exists.

[0094] In the present invention, each radar receiver compensates the estimated parameters to a matched filter detector, and then performs incoherent accumulation of the test statistics of the two receivers and compares them with a detection threshold to achieve target detection.

[0095] Specifically, the delay estimation value, the first Doppler frequency, the second Doppler frequency, and the third Doppler frequency are input into matched filter detectors of multiple radar receivers, and multiple initial test statistics are output, wherein the expression of the initial test statistics is: ;in, For the r The initial test statistic of a radar receiver, tr is the trace of the matrix, For the r The complex matrix signal corresponding to a radar receiver is , For the r The delay estimate corresponding to each radar receiver is: For the rThe invention relates to a method for detecting a target detection object according to the present invention. The method comprises the following steps: determining a first Doppler frequency, a second Doppler frequency and a third Doppler frequency corresponding to each radar receiver; integrating multiple initial test statistics output by multiple radar receivers by using a non-coherent accumulation method to obtain a cumulative test statistic; comparing the cumulative test statistic with a preset detection threshold, and determining that a target detection object exists when the cumulative test statistic is greater than the preset detection threshold.

[0096] In view of the known transmitted signal, a matched filter can be used to maximize the output signal-to-noise ratio of a specific signal, and the estimated parameters can be brought into a matched filter detector to verify the detection performance. The present invention inputs the Doppler frequency estimated by Fourier transform of the first method, the Doppler frequency estimated by fine search of the second method, and the Doppler frequency estimated by multiple signal classification algorithm of the third method into the matched filter detector to obtain multiple initial test statistics.

[0097] For r A radar receiver, its initial test statistic is: ,in , , For the r The delay estimate corresponding to each radar receiver is: For the r The radar receivers correspond to the first Doppler frequency, the second Doppler frequency and the third Doppler frequency.

[0098] The present invention uses a non-coherent accumulation method to integrate data from different radar receivers, which can integrate information from different perspectives and effectively improve detection performance. The present invention involves two receivers. After the two receivers perform parameter estimation respectively, the test statistic of the first receiver is obtained. , the test statistic for the second receiver , so the test statistic processed by the incoherent accumulation method becomes: , the test is: ,in is the detection threshold, is the detection process that compares the test statistic to the detection threshold.

[0099] Experimental verification:

[0100] The present invention considers the situation of 1 radar transmitter, 1 moving target, and 2 radar receivers, assuming that the transmission signal is known. The parameters are set as follows: the number of continuous transmission pulses K=32, the complex Gaussian white noise obeys the complex Gaussian distribution with a mean of 0 and a variance of 1, and the transmission frequency is Hz, since the distributed radar receivers are placed at different distances and angles, the delay, Doppler frequency and reflection coefficient are different. For the first receiver, set the Doppler frequency to Hz, delay , the reflection coefficient amplitude changes with the signal-to-noise ratio, and the phase Randomly generated, , For the second receiver, set the Doppler frequency to Hz, delay , the reflection coefficient amplitude varies with the signal-to-noise ratio, , phase is randomly generated, , To improve the signal-to-noise ratio, the search resolution was set to 0.0001.

[0101] First, the present invention studies the process of estimating the delay and Doppler frequency parameters using one-bit data. The results are from 10 5 Monte Carlo experiment. The first receiver is assumed to sample 32 points of useful signals per pulse, and the useful signal duty cycle is 1%. The second receiver is assumed to sample 64 points of useful signals per pulse, and the useful signal duty cycle is 10%. The present invention uses matched filtering technology to estimate the delay. Figure 2 As shown, Figure 2 is the relationship between the mean absolute error of delay estimation and the signal-to-noise ratio (where, Figure 2 a in is 32 useful signal points, and the duty cycle is 1%. Figure 2 b in the figure is 64 useful signal points and the duty cycle is 10%). As the signal-to-noise ratio increases, the mean absolute error decreases gradually, which indicates that the system can estimate the time delay more accurately under high signal-to-noise ratio conditions. In addition, when the number of sampling points and the duty cycle of the useful signal are higher, the mean absolute error becomes smaller and approaches zero faster. This shows that under the same signal-to-noise ratio conditions, more sampling points and a higher duty cycle can provide more accurate time delay estimation.

[0102] like Figure 3 As shown, Figure 3 is the relationship between the mean square error of Doppler frequency estimation and the signal-to-noise ratio (where, Figure 3 a in is 32 useful signal points, and the duty cycle is 1%. Figure 3 b in the figure represents 64 useful signal points and the duty cycle is 10%), showing the mean square error performance of three different Doppler frequency estimation methods (FFT, FFT fine search and MUSIC) under different signal-to-noise ratios. Obviously, the mean square error of the three methods decreases with the increase of the signal-to-noise ratio. Figure 3In the results, the performance of the FFT fine search method is comparable to that of the FFT method at low SNR. However, when the SNR reaches -10dB, the mean square error of the FFT algorithm tends to stabilize. This stability may be due to the limitation of the number of pulses and sampling points, which makes the frequency resolution of the signal reach its upper limit at higher SNR. In contrast, the mean square error of the FFT fine search method is significantly lower than that of the FFT algorithm, indicating that more useful signal points and higher duty cycle help improve the estimation accuracy. Figure 3 Figure (b) shows that the mean square error of the MUSIC algorithm is relatively high at low signal-to-noise ratios, but decreases significantly as the signal-to-noise ratio increases. Figure 3 The a in the figure shows that the estimation accuracy of the MUSIC algorithm is better than that of the other two methods at low signal-to-noise ratios, and its mean square error also decreases significantly with the increase of the signal-to-noise ratio. Figure 3 The mean square error performance of a in is better than Figure 3 This may be due to the influence of delay estimation coupling on the performance of the MUSIC algorithm. Figure 3 It can be concluded that different Doppler frequency estimation methods show different advantages and disadvantages under different SNR conditions. The FFT fine search method is better than other methods in most cases, while the performance of the MUSIC algorithm is close to that of the FFT fine search method at high SNR.

[0103] Secondly, each radar receiver uses a matched filter detector to detect and extract the target signal and applies delay and Doppler frequency compensation. The matched filter outputs of the two radar receivers are then incoherently accumulated. The present invention compares the performance of this compensated detector with that of a detector without delay and Doppler frequency compensation through the receiver operating characteristic curve. The results are from 10 6 Monte Carlo experiments and show that Figure 4 middle, Figure 4 The receiver operating characteristic curve is: 32 useful signal points, duty cycle is 1% (where, Figure 4 a in the equation is SNR=-16dB, Figure 4 b in the equation is SNR=-14dB). Figure 4 As shown in Figure 2, the detection performance of Doppler compensation using the fine search FFT estimation method is the best, followed by the FFT method. Figure 4 It can be clearly seen that the detection performance of all three methods improves when the SNR increases from -16 dB to -14 dB. In addition, implementing Doppler frequency compensation significantly enhances the detection performance compared to the case without compensation. This shows that the detection performance of the detector is positively correlated with the accuracy of the Doppler frequency estimation. Therefore, accurate parameter estimation is crucial to the performance of the detector.

[0104] The present invention overcomes the limitations of the prior art by studying the single-bit parameter estimation method, realizes robust multi-parameter joint estimation, provides precise support for target detection, improves detector performance through parameter compensation, and reveals the positive correlation between parameter estimation accuracy and target detection effect.

[0105] Furthermore, if Figure 5 As shown, based on the above-mentioned distributed radar target detection method based on single-bit data, the present invention also provides a distributed radar target detection system based on single-bit data, wherein the distributed radar target detection system based on single-bit data includes:

[0106] The single-bit quantization processing module 51 is used to obtain a target received signal and perform single-bit quantization processing on the target received signal to obtain a single-bit digital signal;

[0107] A pulse compression signal generating module 52, configured to perform pulse division processing and matched filtering processing on the single-bit digital signal to obtain a pulse compression signal;

[0108] The delay and Doppler frequency acquisition module 53 is used to acquire a column of data with the largest power and a delay estimation value in the pulse compression signal, and perform Doppler frequency estimation processing according to the column of data with the largest power to obtain multiple Doppler frequency estimation values;

[0109] The target detection object judgment module 54 is used to perform parameter compensation processing according to the delay estimation value and the multiple Doppler frequency estimation values ​​to obtain a cumulative test statistic. When the cumulative test statistic is greater than a preset detection threshold, it is determined that the target detection object exists.

[0110] Furthermore, if Figure 6 As shown, based on the above-mentioned distributed radar target detection method and system based on single-bit data, the present invention also provides a terminal accordingly, and the terminal includes a processor 10, a memory 20 and a display 30. Figure 6 Only some components of the terminal are shown, but it should be understood that it is not required to implement all of the components shown, and more or fewer components may be implemented instead.

[0111] In some embodiments, the memory 20 may be an internal storage unit of the terminal, such as a hard disk or memory of the terminal. In other embodiments, the memory 20 may also be an external storage device of the terminal, such as a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (SecureDigital, SD) card, a flash card (Flash Card), etc. equipped on the terminal. Further, the memory 20 may also include both an internal storage unit of the terminal and an external storage device. The memory 20 is used to store application software and various types of data installed on the terminal, such as the program code of the installation terminal. The memory 20 may also be used to temporarily store data that has been output or is to be output. In one embodiment, a distributed radar target detection program 40 based on single-bit data is stored on the memory 20, and the distributed radar target detection program 40 based on single-bit data can be executed by the processor 10, thereby realizing the distributed radar target detection method based on single-bit data in the present application.

[0112] In some embodiments, the processor 10 may be a central processing unit (CPU), a microprocessor or other data processing chip, used to run the program code or process data stored in the memory 20, such as executing the distributed radar target detection method based on single-bit data.

[0113] In some embodiments, the display 30 may be an LED display, a liquid crystal display, a touch-sensitive liquid crystal display, an OLED (Organic Light-Emitting Diode) touch device, etc. The display 30 is used to display information on the terminal and to display a visual user interface. The components 10-30 of the terminal communicate with each other via a system bus.

[0114] In one embodiment, when the processor 10 executes the distributed radar target detection program 40 based on single-bit data in the memory 20, the steps of the distributed radar target detection method based on single-bit data as described above are implemented.

[0115] The present invention also provides a computer-readable storage medium, wherein the computer-readable storage medium stores a distributed radar target detection program based on single-bit data, and when the distributed radar target detection program based on single-bit data is executed by a processor, the steps of the distributed radar target detection method based on single-bit data as described above are implemented.

[0116] In summary, the present invention provides a distributed radar target detection method, system and terminal based on single-bit data, the method comprising: obtaining a target received signal, and performing single-bit quantization processing on the target received signal to obtain a single-bit digital signal; performing pulse division processing and matched filtering processing on the single-bit digital signal to obtain a pulse compression signal; obtaining a column of data with the largest power and a time delay estimation value in the pulse compression signal, and performing Doppler frequency estimation processing based on the column of data with the largest power to obtain multiple Doppler frequency estimation values; performing parameter compensation processing based on the time delay estimation value and the multiple Doppler frequency estimation values ​​to obtain a cumulative test statistic, and when the cumulative test statistic is greater than a preset detection threshold, it is determined that the target detection object exists. The present invention performs pulse division processing and matched filtering processing on the single-bit digital signal obtained after single-bit quantization processing to determine the maximum power data and the maximum power position, and further calculates the time delay and Doppler frequency, which not only reduces the burden of distributed radar data transmission, but also solves the accuracy problem of parameter recovery of single-bit technology under low signal-to-noise ratio conditions. At the same time, more reliable target detection tasks are achieved through parameter compensation of time delay and Doppler frequency.

[0117] It should be noted that, in this article, the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, article or terminal including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or terminal. In the absence of further restrictions, an element defined by the sentence "comprises a ..." does not exclude the presence of other identical elements in the process, method, article or terminal including the element.

[0118] Of course, those skilled in the art can understand that all or part of the processes in the above-mentioned embodiments can be implemented by instructing related hardware (such as a processor, a controller, etc.) through a computer program, and the program can be stored in a computer-readable storage medium that can be read by a computer, and the program can include the processes of the above-mentioned method embodiments when executed. The computer-readable storage medium can be a memory, a disk, an optical disk, etc.

[0119] It should be understood that the application of the present invention is not limited to the above examples. For ordinary technicians in this field, improvements or changes can be made based on the above description. All these improvements and changes should fall within the scope of protection of the claims attached to the present invention.

Claims

1. A distributed radar target detection method based on single-bit data, characterized in that: The distributed radar target detection method based on single-bit data includes: Acquire a target received signal, and perform single-bit quantization processing on the target received signal to obtain a single-bit digital signal; The single-bit digital signal is subjected to pulse division processing and matched filtering processing to obtain a pulse compression signal; Acquire a column of data with the largest power and a time delay estimation value in the pulse compression signal, and perform Doppler frequency estimation processing according to the column of data with the largest power to obtain multiple Doppler frequency estimation values; The step of acquiring a column of data with the largest power and a time delay estimation value in the pulse compression signal, and performing Doppler frequency estimation processing according to the column of data with the largest power to obtain multiple Doppler frequency estimation values ​​specifically includes: Find a column of data with the largest power and the corresponding maximum power position in the pulse compression signal, and use the maximum power position as the delay estimation value, wherein the expression of the maximum power position is: ； in, is the maximum power position, Solving for the maximum power position, , , is the power data of each column in the pulse compression signal, For the column vector, is the square of the vector's second norm, is the sampling length; Perform Fourier transform processing on the column of data with the largest power, perform spectrum peak search in the frequency domain, obtain a first spectrum peak position, and obtain a first Doppler frequency according to the first spectrum peak position, wherein the expression of the first Doppler frequency is: ； in, is the first Doppler frequency, is the position of the first peak, is the number of pulse signals transmitted; Performing fine frequency compensation processing on the first spectrum peak position to obtain a second Doppler frequency; Constructing a sample covariance matrix according to the target received signal and the time delay estimation value, and performing Doppler frequency estimation processing on the sample covariance matrix using a multiple signal classification algorithm to obtain a third Doppler frequency; The performing fine frequency compensation processing on the first spectrum peak position to obtain the second Doppler frequency specifically includes: The first spectrum peak position is subjected to fine frequency compensation processing by using a fine search Fourier transform algorithm to obtain a second spectrum peak position, wherein the expression of the second spectrum peak position is: ； in, is the second peak position, Solve for the position of the second peak, is the corresponding rotation phase, is the standard Fourier transform matrix OK, is the phase rotation of the original vector, is the column vector corresponding to the maximum power index; A second Doppler frequency is obtained according to the second spectrum peak position, wherein the expression of the second Doppler frequency is: ； in, is the second Doppler frequency; Refining the frequency points near the first spectrum peak position and fine-tuning the peak position to maximize the peak value, thereby overcoming the resolution limitation of Fourier transform; The step of constructing a sample covariance matrix according to the target received signal and the delay estimation value, and performing Doppler frequency estimation processing on the sample covariance matrix using a multiple signal classification algorithm to obtain a third Doppler frequency specifically includes: Constructing a sample covariance matrix according to the target received signal and the delay estimation value, and performing eigenvalue decomposition processing on the sample covariance matrix to obtain a noise subspace; Performing a spectrum search according to the noise subspace to obtain a third spectrum peak position, and obtaining a third Doppler frequency according to the third spectrum peak position; Performing parameter compensation processing according to the time delay estimation value and the multiple Doppler frequency estimation values ​​to obtain a cumulative test statistic, and when the cumulative test statistic is greater than a preset detection threshold, determining that the target detection object exists; The parameter compensation processing is performed according to the time delay estimation value and the multiple Doppler frequency estimation values ​​to obtain a cumulative test statistic, and when the cumulative test statistic is greater than a preset detection threshold, it is determined that the target detection object exists, specifically including: The delay estimation value, the first Doppler frequency, the second Doppler frequency, and the third Doppler frequency are input into matched filter detectors of multiple radar receivers, and multiple initial test statistics are output, wherein the expression of the initial test statistics is: ； in, For the r The initial test statistic of a radar receiver, tr is the trace of the matrix, For the r The complex matrix signal corresponding to a radar receiver is , For the r The delay estimate corresponding to a radar receiver is For the r The first Doppler frequency, the second Doppler frequency and the third Doppler frequency corresponding to the radar receiver, For the r The target receiving signal received by a radar receiver is a single-bit digital signal after single-bit quantization. is the conjugate transpose; Integrating the multiple initial test statistics output by the multiple radar receivers using a non-coherent accumulation method to obtain a cumulative test statistic; The accumulated test statistic is compared with a preset detection threshold, and when the accumulated test statistic is greater than the preset detection threshold, it is determined that the target detection object exists.

2. The distributed radar target detection method based on single-bit data according to claim 1, characterized in that: The step of acquiring a target received signal and performing single-bit quantization processing on the target received signal to obtain a single-bit digital signal specifically includes: A pulse signal is transmitted through a radar transmitter, and a returned target receiving signal is received through a radar receiver, wherein the expression of the target receiving signal is: ； in, For the r The target receiving signal received by the radar receiver is for No. m The target received signal corresponding to each sampling time point is For the r The complex reflection coefficient corresponding to the target reception signal received by a radar receiver is For the quantifier, is the number of pulse signals transmitted, is a single pulse signal, For the r The delay corresponding to the target receiving signal is is the sampling length, is the imaginary unit, For the r The Doppler frequency corresponding to the target reception signal received by a radar receiver, is complex Gaussian white noise, For the m Complex Gaussian white noise corresponding to the sampling time point; The target received signal is subjected to single-bit quantization processing to obtain a single-bit digital signal, wherein the expression of the single-bit digital signal is: ； in, For the r The target receiving signal received by a radar receiver is a single-bit digital signal after single-bit quantization. is the symbolic function, To obtain the real part, To take the imaginary part.

3. The distributed radar target detection method based on single-bit data according to claim 2, characterized in that: The single-bit digital signal is subjected to pulse division processing and matched filtering processing to obtain a pulse compression signal, specifically comprising: Performing pulse division processing on the single-bit digital signal to obtain a received pulse signal; A transmission pulse signal is obtained, and a matched filtering algorithm is used to perform matched filtering on the transmission pulse signal and the received pulse signal to obtain a pulse compression signal, wherein the expression of the pulse compression signal is: ； in, is the pulse compression signal, The first bit of the single-bit digital signal after pulse division processing A pulse signal, is the convolution calculation, is the conjugate transpose.

4. A distributed radar target detection system based on single-bit data, characterized in that: The distributed radar target detection system based on single-bit data includes: A single-bit quantization processing module is used to obtain a target received signal and perform single-bit quantization processing on the target received signal to obtain a single-bit digital signal; A pulse compression signal generating module, used for performing pulse division processing and matched filtering processing on the single-bit digital signal to obtain a pulse compression signal; A time delay and Doppler frequency acquisition module, used to acquire a column of data with the largest power and a time delay estimation value in the pulse compression signal, and perform Doppler frequency estimation processing based on the column of data with the largest power to obtain multiple Doppler frequency estimation values; The step of acquiring a column of data with the largest power and a time delay estimation value in the pulse compression signal, and performing Doppler frequency estimation processing according to the column of data with the largest power to obtain multiple Doppler frequency estimation values ​​specifically includes: Find a column of data with the largest power and the corresponding maximum power position in the pulse compression signal, and use the maximum power position as the delay estimation value, wherein the expression of the maximum power position is: ； in, is the maximum power position, Solving for the maximum power position, , , is the power data of each column in the pulse compression signal, For the column vector, is the square of the vector's second norm, is the sampling length; Perform Fourier transform processing on the column of data with the largest power, perform spectrum peak search in the frequency domain, obtain a first spectrum peak position, and obtain a first Doppler frequency according to the first spectrum peak position, wherein the expression of the first Doppler frequency is: ； in, is the first Doppler frequency, is the position of the first peak, is the number of pulse signals transmitted; Performing fine frequency compensation processing on the first spectrum peak position to obtain a second Doppler frequency; Constructing a sample covariance matrix according to the target received signal and the time delay estimation value, and performing Doppler frequency estimation processing on the sample covariance matrix using a multiple signal classification algorithm to obtain a third Doppler frequency; The performing fine frequency compensation processing on the first spectrum peak position to obtain the second Doppler frequency specifically includes: The first spectrum peak position is subjected to fine frequency compensation processing by using a fine search Fourier transform algorithm to obtain a second spectrum peak position, wherein the expression of the second spectrum peak position is: ； in, is the second peak position, Solve for the position of the second peak, is the corresponding rotation phase, is the standard Fourier transform matrix OK, is the phase rotation of the original vector, is the column vector corresponding to the maximum power index; A second Doppler frequency is obtained according to the second spectrum peak position, wherein the expression of the second Doppler frequency is: ； in, is the second Doppler frequency; Refining the frequency points near the first spectrum peak position and fine-tuning the peak position to maximize the peak value, thereby overcoming the resolution limitation of Fourier transform; The step of constructing a sample covariance matrix according to the target received signal and the delay estimation value, and performing Doppler frequency estimation processing on the sample covariance matrix using a multiple signal classification algorithm to obtain a third Doppler frequency specifically includes: Constructing a sample covariance matrix according to the target received signal and the delay estimation value, and performing eigenvalue decomposition processing on the sample covariance matrix to obtain a noise subspace; Performing a spectrum search according to the noise subspace to obtain a third spectrum peak position, and obtaining a third Doppler frequency according to the third spectrum peak position; Performing parameter compensation processing according to the time delay estimation value and the multiple Doppler frequency estimation values ​​to obtain a cumulative test statistic, and when the cumulative test statistic is greater than a preset detection threshold, determining that the target detection object exists; The parameter compensation processing is performed according to the time delay estimation value and the multiple Doppler frequency estimation values ​​to obtain a cumulative test statistic, and when the cumulative test statistic is greater than a preset detection threshold, it is determined that the target detection object exists, specifically including: The delay estimation value, the first Doppler frequency, the second Doppler frequency, and the third Doppler frequency are input into matched filter detectors of multiple radar receivers, and multiple initial test statistics are output, wherein the expression of the initial test statistics is: ； in, For the r The initial test statistic of a radar receiver, tr is the trace of the matrix, For the r The complex matrix signal corresponding to a radar receiver is , For the r The delay estimate corresponding to a radar receiver is For the r The first Doppler frequency, the second Doppler frequency and the third Doppler frequency corresponding to the radar receiver, For the r The target receiving signal received by a radar receiver is a single-bit digital signal after single-bit quantization. is the conjugate transpose; Integrating the multiple initial test statistics output by the multiple radar receivers using a non-coherent accumulation method to obtain a cumulative test statistic; The accumulated test statistic is compared with a preset detection threshold, and when the accumulated test statistic is greater than the preset detection threshold, it is determined that the target detection object exists.

5. A terminal, characterized in that: The terminal includes: a memory, a processor, and a distributed radar target detection program based on single-bit data stored in the memory and executable on the processor. When the distributed radar target detection program based on single-bit data is executed by the processor, the steps of the distributed radar target detection method based on single-bit data according to any one of claims 1 to 3 are implemented.

6. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a distributed radar target detection program based on single-bit data, and when the distributed radar target detection program based on single-bit data is executed by a processor, the steps of the distributed radar target detection method based on single-bit data as described in any one of claims 1 to 3 are implemented.

Images