Interference detection method based on multi-dimensional change domain conjoint analysis

Abstract

The invention discloses an interference detection method based on multi-dimensional change domain conjoint analysis, which comprises the following steps of: performing Fourier transform on an SAR (Synthetic Aperture Radar) interference-containing single-view complex image along a distance dimension to obtain a representation matrix of the image in a distance frequency domain-azimuth time domain; based on the representation matrix, determining the number of interference-containing pulses, and constructing an interference-containing signal pulse matrix; performing fractional order Fourier transform on the pulse matrix containing the interference signals, and determining the optimal order of the pulses containing the interference signals; based on the relationship between the optimal order of the interference signal-containing pulse and a preset order, determining that the interference signal may be intermittent forwarding interference or linear frequency modulation interference, or the interference signal may be narrowband interference or sine frequency modulation interference; determining intermittent forwarding interference or linear frequency modulation interference by using the interference signal frequency modulation rate and the SAR echo signal frequency modulation rate; and determining narrowband interference or sine frequency modulation interference based on the interference signal bandwidth and the SAR echo signal bandwidth. The method can realize preliminary classification of interference, and is high in detection speed and easy for engineering realization.

Description

Technical Field

[0001] The present invention relates to the field of radar signal processing, and in particular to an interference detection and parameter extraction method based on multi-dimensional transform domain joint analysis. Background Art

[0002] Radar remote sensing satellites are usually equipped with Synthetic Aperture Radar (SAR) as an important active microwave remote sensing detection payload, which can obtain physical characteristic information of ground objects and targets all day and all weather, and obtain two-dimensional high-resolution images of the scene. The low-frequency band, high-resolution radar remote sensing satellite transmits signals with large instantaneous bandwidth, wide beam, and long dwell time, and is easily interfered by other electromagnetic radiation sources in the same frequency band. Electromagnetic interference will affect the imaging coherent focusing process, resulting in blurred and defocused images or the appearance of covering artifacts, so it is necessary to study high-precision and high-efficiency interference suppression methods.

[0003] Interference detection is an important prerequisite for interference suppression. The purpose of interference detection is to determine whether the interference signal exists and to distinguish its interference type, so as to select a more effective interference suppression method and achieve refined interference suppression. However, there is currently no fast and efficient interference detection method for several common types of interference: intermittent forwarding interference, narrowband interference, linear frequency modulation interference, and sinusoidal frequency modulation interference. Summary of the invention

[0004] The present invention designs an interference detection method based on multi-dimensional change domain joint analysis. First, the pulses of the interference signal are detected and extracted in the distance frequency domain. Then, the frequency modulation rates of different types of interference signals are calculated using fractional-order Fourier transform. The bandwidth of the interference signal is estimated by combining the calculation of the kurtosis distribution curve. This method can quickly distinguish between intermittent forwarding interference, linear frequency modulation interference, narrowband interference and sinusoidal frequency modulation interference.

[0005] In order to achieve the above tasks, the present invention adopts the following technical solutions:

[0006] An interference detection method based on multi-dimensional change domain joint analysis, comprising:

[0007] Fourier transform is performed on the SAR single-view complex image containing interference along the range dimension to obtain the image representation matrix in the range frequency domain-azimuth time domain; based on the representation matrix, the number of interference pulses is determined and the interference signal pulse matrix is ​​constructed;

[0008] Performing fractional Fourier transform on the pulse matrix containing interference signals to determine the optimal order of the pulses containing interference signals;

[0009] Based on the relationship between the optimal order of the interference signal pulse and the preset order, it is determined that the interference signal may be intermittent forwarding interference or linear frequency modulation interference, or the interference signal may be narrowband interference or sinusoidal frequency modulation interference;

[0010] When the interference signal may be intermittent forwarding interference or linear frequency modulation interference, the specific type of the interference signal is determined by setting the interference detection error and the interference signal modulation frequency and the SAR echo signal modulation frequency;

[0011] When the interference signal may be narrowband interference or sinusoidal frequency modulation interference, the interference signal bandwidth is estimated, and the specific type of the interference signal is determined based on the interference signal bandwidth and the SAR echo signal bandwidth.

[0012] Furthermore, the method of determining the number of interference pulses based on the characterization matrix and constructing an interference signal pulse matrix includes:

[0013] For size N r ×N a The two-dimensional image matrix x(t r ,t a ) performs Fourier transform along the distance dimension to obtain the image representation matrix X(f r ,t a ), fr represents the distance frequency after Fourier transform; Nr and Na represent the number of sampling points in the distance dimension and the orientation dimension of the image, respectively, r represents the azimuth sampling time, t a represents the distance sampling time;

[0014] For X(f r ,t a ) Perform kurtosis detection along the azimuth dimension to obtain the kurtosis detection curve of the azimuth dimension; By setting the interference threshold and detecting and distinguishing, the number of interference pulses t r_inter , and from the two-dimensional image matrix x(t r ,t a ) extracts the interference signal pulses and constructs the interference signal pulse matrix x(t r ,t r_inter ).

[0015] Further, performing a fractional Fourier transform on the interference signal pulse matrix to determine the optimal order of the interference signal pulses includes:

[0016] For the pulse matrix x(t a ,t r_inter ) performs fractional Fourier transform FRFT, with an interval of 0.01, traverses all orders within p = [0,2], and obtains x(t a ,t r_inter)'s two-dimensional u-domain matrix X p (p,u p ),u p Indicates the FRFT result corresponding to the order p; X p (p,u p ) accumulates energy along the order dimension to obtain the u-domain accumulation curve, and detects the order corresponding to the minimum energy value in the accumulation curve, which is the optimal order p of the pulse containing the interference signal opt .

[0017] Further, the determining, based on the relationship between the optimal order of the interference signal pulse and the preset order, that the interference signal may be intermittent forwarding interference or linear frequency modulation interference, or that the interference signal may be narrowband interference or sinusoidal frequency modulation interference, includes:

[0018] When the optimal order of the interference signal pulse p opt When ≠1, it indicates that the interference signal in the frequency domain cannot achieve the energy aggregation effect. Based on this, it can be judged that the interference signal may be intermittent forwarding interference or linear frequency modulation interference; otherwise, the interference may be narrowband interference or sinusoidal frequency modulation interference.

[0019] Furthermore, when the interference signal may be intermittent forwarding interference or linear frequency modulation interference, the specific type of the interference signal is determined by setting the interference detection error and the frequency modulation of the interference signal and the frequency modulation of the SAR echo signal; including:

[0020] Using the optimal order p opt Solve the interference signal modulation frequency K r_inter , further determine the interference type:

[0021]

[0022] Among them, PRT represents the sampling time interval of the SAR echo signal in the range direction, and F s Indicates the sampling rate in the range direction;

[0023] By comparing the interference signal modulation frequency K r_inter The SAR echo signal is modulated at frequency K r_sar , which can distinguish intermittent forwarding interference and linear frequency modulation interference;

[0024] Set the interference detection error to 0.01. When , the interference type is intermittent forwarding interference; When , the interference is linear frequency modulation interference.

[0025] Furthermore, when the interference signal may be narrowband interference or sinusoidal frequency modulation interference, estimating the interference signal bandwidth, and determining the specific type of the interference signal based on the interference signal bandwidth and the SAR echo signal bandwidth, includes:

[0026] The kurtosis distribution curve is calculated by detecting the distance spectrum of the single-view complex image along the distance dimension, detecting the total number of outliers in the kurtosis distribution curve, that is, the number of distance lines where the interference is located, and then estimating the interference signal bandwidth. The calculation formula is:

[0027]

[0028] when When , the interference is narrowband interference; when When , the interference is sinusoidal frequency modulation interference; the SAR echo signal bandwidth is determined by the radar transmission parameters.

[0029] A terminal device comprises a processor, a memory and a computer program stored in the memory; when the processor is executed by a computer, the interference detection method based on multi-dimensional change domain joint analysis is implemented.

[0030] A computer-readable storage medium stores a computer program; when the computer program is executed by a processor, the interference detection method based on multi-dimensional change domain joint analysis is implemented.

[0031] Compared with the prior art, the present invention has the following design features:

[0032] The present invention innovatively uses fractional Fourier transform to estimate the frequency modulation rate of different types of interference signals, and combines kurtosis to estimate the bandwidth of interference signals, which can quickly distinguish intermittent forwarding interference, linear frequency modulation interference signals, narrowband signals, and sinusoidal frequency modulation broadband interference, which is conducive to the subsequent use of specific interference suppression strategies for different interferences. At the same time, the optimal order of different interference signals is solved by fractional Fourier transform, and it is preliminarily judged whether the interference is an interference signal with linear frequency modulation characteristics. The optimal order is further used to estimate the frequency modulation rate of the interference signal to determine whether the interference is intermittent forwarding interference. Finally, the interference is determined to be narrowband interference or sinusoidal frequency modulation interference by estimating the signal bandwidth. The present invention can realize the preliminary classification of interference, which is convenient for the subsequent selection of corresponding interference suppression methods for different interference types, has fast detection speed, and is easy to implement in engineering. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Figure 1 It is a schematic diagram of the implementation flow in one embodiment of the method of the present invention;

[0034] Figure 2 Time-frequency diagram of SAR echo signals with four different interferences processed by the method of the present invention; (a) is intermittent forwarding interference, (b) is linear frequency modulation interference, (c) is narrowband interference, and (d) is sinusoidal frequency modulation interference;

[0035] Figure 3(a) to (d) correspond to the u-domain accumulation curves of fractional Fourier transform of intermittent forwarding interference, linear frequency modulation interference, narrowband interference and sinusoidal frequency modulation interference data, respectively. The horizontal axis represents the ergodic order p = [0,2], sampled at 0.01, with a total of 2001 sampling points, and the vertical axis represents the u-domain accumulation value at different orders. DETAILED DESCRIPTION

[0036] The basic idea of ​​the present invention is that for intermittent forwarding interference and linear frequency modulation interference, the distance spectrum of the SAR echo signal can be processed by fractional Fourier transform (FRFT), the optimal rotation angle can be calculated, and then the frequency modulation rate of the interference signal can be calculated; while for narrowband interference and sinusoidal frequency modulation interference signals, the signal aggregation effect is best in Fourier transform (FFT domain), that is, the optimal order is 1. Therefore, the type of interference can be preliminarily determined by FRFT. Then, for intermittent forwarding interference and linear frequency modulation interference, the frequency modulation rate of the interference signal can be calculated using the optimal order to achieve further classification; for narrowband interference and sinusoidal frequency modulation interference, the bandwidth of the interference signal is estimated by calculating the distance dimension kurtosis distribution curve in the distance frequency domain to achieve further classification, thereby achieving rapid detection and distinction of four typical interference signals.

[0037] Referring to the accompanying drawings, an interference detection method based on fractional-order Fourier transform provided by the present invention comprises the following steps:

[0038] Step 1: for size N r ×N a The two-dimensional image matrix x(t r ,t a ) to perform processing, where Nr and Na represent the number of sampling points in the image distance dimension and azimuth dimension respectively, tr represents the azimuth sampling time, and t a Represents the distance sampling time.

[0039] First, in the image coordinates, the image is Fourier transformed along the x-axis (distance dimension) to obtain the image representation matrix X(f r ,t a ), fr represents the distance frequency after Fourier transformation; then, for X(f r ,t a ) performs kurtosis detection along the y-axis (azimuth dimension) to obtain the kurtosis detection curve of the azimuth dimension; by setting the interference threshold and detecting and distinguishing, the number of interference pulses t r_inter , and from the two-dimensional image matrix x(t r ,t a ) extracts the interference signal pulses and constructs the interference signal pulse matrix x(t r ,t r_inter), that is, selecting the row containing the interfering pulse in the two-dimensional image matrix and extracting it.

[0040] Step 2: For the interference signal pulse matrix x(t a ,t r_inter ) performs fractional Fourier transform (FRFT), with an interval of 0.01, traverses all orders within p = [0,2], and obtains x(t a ,t r_inter )'s two-dimensional u-domain matrix X p (p,u p ),u p Indicates the FRFT result corresponding to the order p; X p (p,u p ) accumulates energy along the x-axis (order dimension) to obtain the u-domain accumulation curve, and detects the order p corresponding to the minimum energy value in the accumulation curve, and the optimal order p of the pulse containing the interference signal can be obtained opt .

[0041] Step 3: Preliminarily determine whether the interference signal is intermittent forwarding interference or linear frequency modulation interference according to the preset order 1: opt When ≠1, it means that the interference signal in the frequency domain cannot achieve the energy aggregation effect. Based on this, it can be judged that the interference signal may be intermittent forwarding interference or linear frequency modulation interference; otherwise, the interference may be narrowband interference or sinusoidal frequency modulation interference.

[0042] When the optimal order p opt ≠1, the interference signal may be intermittent forwarding interference or linear frequency modulation interference. opt Solve the interference signal modulation frequency K r_inter , further determine the interference type:

[0043]

[0044] Among them, PRT represents the sampling time interval of the SAR echo signal in the range direction, and F s Indicates the sampling rate in the range direction.

[0045] By comparing the interference signal modulation frequency K r_inter The modulation frequency of the SAR echo signal (determined by the radar transmission parameters) K r_sar , which can distinguish intermittent forwarding interference and linear frequency modulation interference;

[0046] Set the interference detection error to 0.01. When , the interference type is intermittent forwarding interference; When , the interference is linear frequency modulation interference.

[0047] Intermittent forwarding interference is defined as the interference signal being highly coherent with the echo signal. The SAR echo signal is a linear frequency modulation signal. In this step, by comparing the frequency modulation rates of the interference signal and the echo signal, it can be determined whether the interference is intermittent forwarding interference.

[0048] Step 4: Sinusoidal FM interference is a type of broadband interference. Narrowband interference and broadband interference can be defined by bandwidth size, so it can be determined by calculating the interference signal bandwidth.

[0049] When the optimal order p opt = 1, the interference may be narrowband interference or sinusoidal frequency modulation interference, which can be detected by detecting the range spectrum X(f) of the single-view complex image containing interference along the x-axis (distance dimension). a ,t r ) calculate the kurtosis distribution curve, detect the total number of outliers in the kurtosis distribution curve, that is, the number of interference distance lines; f a is the signal frequency; then the interference signal bandwidth is estimated using the following formula:

[0050]

[0051] when When , the interference is narrowband interference; when When , the interference is sinusoidal frequency modulation interference; the SAR echo signal bandwidth is determined by the radar transmission parameters.

[0052] Example:

[0053] Step 1: Perform Fourier transform on the SAR interference SLC image with an image size of 2336×3398 along the range direction to obtain the representation matrix X(f a ,t r ). Then, for X(f a ,t r ) Perform kurtosis detection along the distance frequency domain to obtain a kurtosis detection curve. By setting the interference threshold, the number of interference pulses t r_inter , and extract the interference signal pulse x(2336,t r_inter ), size is 1×2336.

[0054] Step 2: For x(2336,t r_inter ) performs fractional Fourier transform (FRFT), takes 0.01 as sampling interval, and traverses all orders in [0,2] to obtain x(2336,t r_inter ) of the two-dimensional u-domain matrix X p (p,u p ). p (p,u p) is used to accumulate energy along the order dimension to obtain the u-domain accumulation curve. By detecting the minimum value, the interference signal pulse x(t a ,t r_inter )Optimal order p opt =1.646.

[0055] Figure 3 The optimal order of intermittent forwarding interference p is shown in (a) opt =1.646, (b) shows the optimal order p of linear frequency modulation interference opt =1.424, (c) shows the optimal order p of narrowband interference opt =1, (b) shows the optimal order p of sinusoidal FM interference opt =1.

[0056] Step 3, when the optimal order p opt ≠1, using the optimal order p opt Solve the interference signal modulation frequency and find the interference signal modulation frequency K r_inter =-7.2192×10 11 .

[0057] again Therefore, this interference type is intermittent forwarding interference.

[0058] Table 1 SAR echo signal parameters

[0059]

[0060] The above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. These modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application, and should all be included in the protection scope of the present application.

Claims

1. An interference detection method based on multi-dimensional change domain joint analysis, characterized in that: include: Fourier transform is performed on the SAR single-view complex image containing interference along the range dimension to obtain the image representation matrix in the range frequency domain and azimuth time domain. Based on the characterization matrix, the number of pulses containing interference is determined, and a pulse matrix containing interference signals is constructed; Perform fractional Fourier transform on the pulse matrix containing interference signals to determine the optimal order of the pulses containing interference signals; Based on the relationship between the optimal order of the interference signal pulse and the preset order, it is determined that the interference signal may be intermittent forwarding interference or linear frequency modulation interference, or the interference signal may be narrowband interference or sinusoidal frequency modulation interference; When the interference signal may be intermittent forwarding interference or linear frequency modulation interference, the specific type of the interference signal is determined by setting the interference detection error and the interference signal modulation frequency and the SAR echo signal modulation frequency; When the interference signal may be narrowband interference or sinusoidal frequency modulation interference, the interference signal bandwidth is estimated, and the specific type of the interference signal is determined based on the interference signal bandwidth and the SAR echo signal bandwidth.

2. The interference detection method based on multi-dimensional change domain joint analysis according to claim 1 is characterized in that: The method of determining the number of interference pulses based on the characterization matrix and constructing an interference signal pulse matrix includes: For size N r ×N a The two-dimensional image matrix x(t r ,t a ) performs Fourier transform along the distance dimension to obtain the image representation matrix X(f r ,t a ), fr represents the distance frequency after Fourier transform; Nr and Na represent the number of sampling points in the distance dimension and the orientation dimension of the image, respectively, r represents the azimuth sampling time, t a represents the distance sampling time; For X(f r ,t a ) Perform kurtosis detection along the azimuth dimension to obtain the kurtosis detection curve of the azimuth dimension; By setting the interference threshold and detecting and distinguishing, the number of interference pulses t r_inter , and from the two-dimensional image matrix x(t r ,t a ) extracts the interference signal pulses and constructs the interference signal pulse matrix x(t r ,t r_inter ).

3. The interference detection method based on multi-dimensional change domain joint analysis according to claim 1 is characterized in that: The step of performing a fractional Fourier transform on the interference signal pulse matrix to determine the optimal order of the interference signal pulses includes: For the pulse matrix x(t a ,t r_inter ) performs fractional Fourier transform FRFT, with an interval of 0.01, traverses all orders within p = [0,2], and obtains x(t a ,t r_inter )'s two-dimensional u-domain matrix X p (p,u p ),u p Indicates the FRFT result corresponding to the order p; X p (p,u p ) accumulates energy along the order dimension to obtain the u-domain accumulation curve, and detects the order corresponding to the minimum energy value in the accumulation curve, which is the optimal order p of the pulse containing the interference signal opt .

4. The interference detection method based on multi-dimensional change domain joint analysis according to claim 1 is characterized in that: Based on the relationship between the optimal order of the interference signal pulse and the preset order, it is determined that the interference signal may be intermittent forwarding interference or linear frequency modulation interference, or the interference signal may be narrowband interference or sinusoidal frequency modulation interference; include: When the optimal order of the interference signal pulse p opt When ≠1, it indicates that the interference signal in the frequency domain cannot achieve the energy aggregation effect. Based on this, it can be judged that the interference signal may be intermittent forwarding interference or linear frequency modulation interference; otherwise, the interference may be narrowband interference or sinusoidal frequency modulation interference.

5. The interference detection method based on multi-dimensional change domain joint analysis according to claim 1 is characterized in that: When the interference signal may be intermittent forwarding interference or linear frequency modulation interference, the specific type of the interference signal is determined by setting the interference detection error and the interference signal modulation frequency and the SAR echo signal modulation frequency; including: Using the optimal order p opt Solve the interference signal modulation frequency K r_inter , further determine the interference type: Among them, PRT represents the sampling time interval of the SAR echo signal in the range direction, and F s Indicates the sampling rate in the range direction; By comparing the interference signal modulation frequency K r_inter The SAR echo signal modulation frequency K r_sar , which can distinguish intermittent forwarding interference and linear frequency modulation interference; Set the interference detection error to 0.

01. When , the interference type is intermittent forwarding interference; When , the interference is linear frequency modulation interference.

6. The interference detection method based on multi-dimensional change domain joint analysis according to claim 1 is characterized in that: When the interference signal may be narrowband interference or sinusoidal frequency modulation interference, estimating the interference signal bandwidth, and determining the specific type of the interference signal based on the interference signal bandwidth and the SAR echo signal bandwidth, includes: The kurtosis distribution curve is calculated by detecting the distance spectrum of the single-view complex image along the distance dimension, detecting the total number of outliers in the kurtosis distribution curve, that is, the number of distance lines where the interference is located, and then estimating the interference signal bandwidth. The calculation formula is: when When , the interference is narrowband interference; when When , the interference is sinusoidal frequency modulation interference; the SAR echo signal bandwidth is determined by the radar transmission parameters.

7. A terminal device, comprising a processor, a memory and a computer program stored in the memory; characterized in that: When the processor is executed by a computer, the interference detection method based on multi-dimensional change domain joint analysis according to any one of claims 1 to 6 is implemented.

8. A computer-readable storage medium, wherein a computer program is stored in the medium; characterized in that: When the computer program is executed by a processor, the interference detection method based on multi-dimensional change domain joint analysis according to any one of claims 1 to 6 is implemented.

Images