A radar signal distance alignment implementation method based on envelope correlation

Abstract

The present application relates to a kind of envelope correlation-based radar signal distance alignment implementation method, belong to radar signal processing technical field.It includes: the original signal envelope of each pulse under the same pulse group is calculated;Envelope cross-correlation between original signal envelope is calculated;The peak position of envelope cross-correlation of original signal is obtained;Original signal is N times interpolation, and the envelope of signal after interpolation is calculated;The peak position of envelope cross-correlation of original signal is obtained in the corresponding position after interpolation, and the envelope cross-correlation in the range of N / 2+1 sampling points in the vicinity of left and right around with the position as center is calculated;The peak position of envelope cross-correlation of signal after interpolation is obtained;According to the peak position of envelope cross-correlation of signal after interpolation, signal after interpolation is compensated;Compensated signal is N times decimation.It solves the problem that processing performance is low due to excessive calculation in the envelope correlation-based wideband radar signal distance alignment processing.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of radar signal processing, and particularly relates to distance alignment processing of radar wideband signals, and in particular to a radar signal distance alignment efficient implementation method based on envelope correlation. BACKGROUND

[0002] One of the functions of wideband signals is to provide a good basis for radar target recognition. Conventional narrowband radars have very low range resolution, and general targets (such as aircraft) appear as "point" targets. Although the waveforms also contain certain target information, they are very rough. The target echo of a radar with a frequency bandwidth of more than 100 MHz to several hundred MHz is a high range resolution (HRR) signal, and the resolution can reach sub-millimeter. The HRR echo signal of a general target appears as a one-dimensional range image. However, the high range resolution of wideband signals also brings some inconvenience in application, mainly that the distance migration of a target echo between pulses easily causes the distance cells of the image to be misaligned, and it is difficult to realize coherent accumulation of a string of echoes.

[0003] For the distance cell migration of a high-resolution range image, under certain conditions, algorithms can be used for remediation. Common algorithms include envelope correlation method, overall correlation method, and minimum entropy method. Among them, the envelope correlation method and the overall correlation method both belong to cross-correlation methods, and are aligned according to the maximum of the real envelope cross-correlation of the echo. The minimum entropy method is aligned according to the minimum entropy of the real envelope of the echo. The envelope correlation method uses the cross-correlation method on the real envelopes of two adjacent echoes to take the time delay corresponding to the peak value as compensation. In fact, the radar signal is recorded by discrete time sampling, and the sampling interval is generally slightly smaller than the pulse width. However, the envelope alignment accuracy requirement is 1 / 8 of the distance cell, so when calculating the cross-correlation function, the time discrete value is usually interpolated by 8 times. In this way, the adjacent real envelopes can be well aligned, and the algorithm operation amount is undoubtedly greatly improved, which is difficult to apply in engineering. SUMMARY

[0004] The technical problem to be solved by the present application is:

[0005] In order to avoid the shortcomings of the prior art, the present application provides a radar signal distance alignment efficient implementation method based on envelope correlation, which is used to solve the problem of large operation amount and insufficient processing performance when using the envelope correlation method to perform distance alignment on radar wideband signals.

[0006] In order to solve the above technical problems, the technical scheme adopted by the present application is:

[0007] A radar signal distance alignment efficient implementation method based on envelope correlation, characterized in that it comprises:

[0008] Calculate the original signal envelope of each pulse under the same pulse group;

[0009] Select the first pulse as the reference pulse, and calculate the envelope cross-correlation between the remaining pulse envelopes and the reference pulse envelope;

[0010] Obtain the peak position of the envelope cross-correlation of the original signal;

[0011] Perform N-fold interpolation on the original signal and calculate the envelope of the interpolated signal;

[0012] The peak position of the envelope cross-correlation of the original signal is obtained at the corresponding position after interpolation, and the envelope cross-correlation is calculated within a range of N / 2+1 sampling points to the left and right of that position as the center.

[0013] Obtain the peak position of the envelope cross-correlation of the interpolated signal;

[0014] The interpolated signal is compensated based on the peak position of the envelope cross-correlation of the interpolated signal;

[0015] The compensated signal is decimated by a factor of N.

[0016] A further technical solution of the present invention: the calculation of the envelope cross-correlation between the residual pulse envelope and the reference pulse envelope includes:

[0017] Perform 2N operations on the reference pulse envelope u1(n) r The -1 point Fourier transform yields U1(k), k∈[1,2N]. r -1];

[0018] For the i-th remaining pulse envelope u i (n), i∈[2,N) d Do 2N r -1 point Fourier transform yields U i (k), k∈[1,2N] r -1];

[0019] For U1 and U i The result of the dot product of the conjugate vectors is taken as an inverse Fourier transform to obtain the envelope cross-correlation result R. 1i ,i∈[2,N d ].

[0020] A further technical solution of the present invention: the acquisition of the peak position of the envelope cross-correlation of the original signal specifically involves:

[0021] By iterating through the envelope cross-correlation R of the i-th remaining pulse and the reference pulse 1i Obtain its peak position a i ,a i ∈[-N r +1,N r -1], i∈[2,N d ].

[0022] The further technical scheme of the present application is that the original signal is N times interpolated, and the envelope of the interpolated signal is calculated, specifically:

[0023] The original signal s i (n), n ∈ [1, N r ], i ∈ [1, N d ] is Fourier transformed to obtain S i (k), k ∈ [1, N r ], i ∈ [1, N d ];

[0024] N.N r zeros are continuously inserted from ceil(N r / 2+1) of S i (k) to obtain S i '(k'), k' ∈ [1, (N+1)N r ];

[0025] S'(k') is inverse Fourier transformed to obtain the interpolated signal s'(n'), n' ∈ [1, (N+1)N r ];

[0026] The modulus of s'(n') is taken to obtain the envelope u'(n') of the interpolated signal, n' ∈ [1, (N+1)N r ].

[0027] The further technical scheme of the present application is that the peak position of the envelope cross-correlation of the original signal is obtained, the corresponding position after interpolation is calculated, and the envelope cross-correlation in the range of N / 2+1 sampling points near the left and right center is calculated, specifically:

[0028] The peak position a i of the envelope cross-correlation R 1i , the corresponding position after interpolation is a' i =(N+1)a i , i ∈ [2, N d ];

[0029] The envelope cross-correlation result R' 1i , i ∈ [2, N d ] in the range of N / 2+1 sampling points near the left and right center with a' i as the center is obtained by the following formula

[0030] R' 1i (m) = ∑u'1(n')u' i (n'-m), m ∈ [a' i -N / 2-1, a' i +N / 2+1]

[0031] Wherein, u'1 is the envelope of the reference pulse after interpolation, u' i is the envelope of the i-th pulse after interpolation.

[0032] The further technical solution of the present application is that the compensated signal after interpolation is compensated according to the peak position of the envelope cross-correlation of the signal after interpolation, specifically:

[0033] The peak position b 1i of the envelope cross-correlation R i of the i-th pulse signal after interpolation is obtained by the following formula: i The frequency domain result X

[0034]

[0035] Wherein, S i (k) is the frequency domain result of the i-th pulse signal before compensation, i∈[2,N d ]。

[0036] The further technical solution of the present application is that the compensated signal after interpolation is compensated according to the peak position of the envelope cross-correlation of the signal after interpolation, specifically:

[0037] The first ceil(N i / 2) data and the last N r -ceil(N r / 2) data of X r (k') are combined to form X i (k), k∈[1,N r ], i∈[2,N d ];

[0038] The inverse Fourier transform is performed on X i (k) to obtain the distance-aligned signal x i (n), n∈[1,N r ], wherein N r is the number of sampling points, and ceil(.) represents rounding up.

[0039] A computer system, characterized in comprising: one or more processors, a computer readable storage medium for storing one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors implement the above method.

[0040] A computer readable storage medium, characterized in storing computer executable instructions, the instructions being executed to implement the above method.

[0041] A computer program product characterized in comprising computer executable instructions for implementing the above method when executed.

[0042] The present application has the advantages of:

[0043] The present application provides a radar signal distance alignment efficient implementation method based on envelope correlation, which optimizes the cross-correlation operation and compensation operation to improve the algorithm processing performance, including: interpolation, correlation, compensation and decimation operation of the signal in the frequency domain, using efficient scientific computing library for acceleration, such as FFTW, OpenBlas library, etc.; optimizing the correlation operation to avoid the complete length envelope cross-correlation operation after interpolation; optimizing the compensation operation to avoid the complete compensation coefficient calculation after interpolation, and only compensating for the sampling points required for final decimation. Since the method of the present application avoids the complete envelope cross-correlation and compensation of the interpolated signal, the computational complexity of the entire algorithm is greatly reduced, thereby improving the processing performance of the algorithm. BRIEF DESCRIPTION OF DRAWINGS

[0044] The accompanying drawings are included to provide a further understanding of the application and are incorporated in and constitute a part of this specification, illustrate embodiments of the application and together with the description serve to explain the principles of the application. In the drawings:

[0045] Figure 1 The flowchart is used to implement the present application.

[0046] Figure 2 The signal envelope before distance alignment processing.

[0047] Figure 3 The signal envelope after processing by the distance alignment method of the present application. DETAILED DESCRIPTION

[0048] In order to make the purpose, technical scheme and advantages of the present application clearer, the present application will be further described in detail below in combination with the drawings and examples. It should be understood that the specific embodiments described herein are only used to explain the present application and do not limit the present application. In addition, the technical features involved in each embodiment of the present application described below can be combined with each other as long as they do not conflict with each other.

[0049] Reference Figure 1 The present embodiment provides a radar signal distance alignment efficient implementation method based on envelope correlation, comprising the following steps:

[0050] S101, calculate the original signal envelope of each pulse under the same pulse group, and select the first pulse as the reference pulse;

[0051] S102, calculate the envelope cross-correlation between the remaining pulses and the reference pulse;

[0052] S103, Obtain the peak position of the envelope cross-correlation;

[0053] S104, perform N-fold interpolation on the original signal and calculate the envelope of the interpolated signal;

[0054] S105: Obtain the result of S103 at the corresponding position after interpolation, and calculate the envelope cross-correlation within a range of N / 2+1 sampling points to the left and right of that position as the center.

[0055] S106, Obtain the peak position of the interpolated envelope cross-correlation;

[0056] S107, compensate the interpolated signal based on the peak position of S106.

[0057] The above steps are implemented as follows:

[0058] In step S101, by adjusting each pulse s under the pulse group i (n), n∈[1,N r ], i∈[1,N d Perform modulo operation to obtain the original signal envelope u. i (n), n∈[1,N r ], i∈[1,N d ], where N r N is the number of sampling points. d The first pulse is selected as the reference pulse, representing the number of pulses in the pulse group.

[0059] In step S102, the envelope cross-correlation between the remaining pulse and the reference pulse is calculated. The specific implementation steps are as follows:

[0060] 2a) Perform 2N operations on the reference pulse envelope u1(n) r The -1 point Fourier transform yields U1(k), k∈[1,2N]. r -1];

[0061] 2b) For the envelope u of the i-th pulse i (n), i∈[2,N) d Do 2N r -1 point Fourier transform yields U i (k), k∈[1,2N] r -1];

[0062] 2c) For U1 and U i The result of the dot product of the conjugate vectors is taken as an inverse Fourier transform to obtain the envelope cross-correlation result R. 1i ,i∈[2,N d ];

[0063] In step S103, the envelope cross-correlation R 1i is obtained by traversing the envelope of the i-th pulse and the reference pulse i , i ∈[-N r +1,N r -1],i∈[2,N d ];

[0064] In step S104, the original signal is N-fold interpolated, and the envelope of the interpolated signal is calculated. The specific implementation steps are as follows:

[0065] 4a) Fourier transform the original signal s i (n),n∈[1,N r ],i∈[1,N d ] to obtain S i (k),k∈[1,N r ],i∈[1,N d ];

[0066] 4b) From S i (k) ceil(N r / 2+1) continuously insert N.N r zeros to obtain S i '(k'),k'∈[1,(N+1)N r ];

[0067] 4c) Inverse Fourier transform S'(k') to obtain the interpolated signal s'(n'),n'∈[1,(N+1)N r ];

[0068] 4d) Modulus s'(n') to obtain the interpolated signal envelope u'(n'),n'∈[1,(N+1)N r ];

[0069] In step S105, the envelope cross-correlation corresponding position of the result of step S103 after interpolation is obtained, and the envelope cross-correlation is calculated in the range of N / 2+1 sampling points around the center position. The specific implementation steps are as follows:

[0070] 5a) The peak position a 1i of the envelope cross-correlation R i obtained in step (3) is obtained, and the corresponding position after interpolation is

[0071] a' i =(N+1)a i ,i∈[2,N d ];

[0072] 5b) a' iThe envelope cross-correlation result R' of the sampled area within N / 2+1 sampling points to the left and right of the center. 1i ,i∈[2,N d The following formula is used to obtain...

[0073] R' 1i (m)=∑u'1(n')u' i (n'-m), m∈[a' i -N / 2-1,a' i +N / 2+1]

[0074] Where u'1 is the envelope after interpolation of the reference pulse, u' i The envelope after interpolation of the i-th pulse;

[0075] In step S106, the envelope cross-correlation R is calculated by traversing the interpolated values ​​of the i-th pulse and the reference pulse. 1i 'Get its peak position b' i ,b i ∈[-(N+1)N r +1,(N+1)N r -1], i∈[2,N d ];

[0076] In step S107, the interpolated signal is compensated based on the peak position from step S106. The specific implementation steps are as follows:

[0077] The envelope cross-correlation R can be obtained through step S106. 1i 'peak position b' i Then the frequency domain result X of the i-th interpolated pulse signal after compensation i The value of ' can be obtained from the following formula.

[0078]

[0079] Among them, S i '(k) represents the frequency domain result of the i-th interpolated pulse signal before compensation, i∈[2,N] d ];

[0080] In step S108, the compensated signal is decimated by a factor of N. The specific implementation steps are as follows:

[0081] 8a) Extract X i The first ceil (N) of '(k') r / 2) data points and the last N r -ceil(N r / 2) data are combined to form X i (k), k∈[1,N] r ], i∈[2,N d ];

[0082] 8b) to X i (k) inverse Fourier transform to get the range aligned signal x i (n), n∈[1, N r ]; wherein, N r is the number of sampling points, N is the interpolation multiple, and ceil(.) represents rounding up.

[0083] The effects of the present application are further illustrated by the following comparative tests:

[0084] 1. Experimental parameters

[0085] The CPU model used by the server is Feng 2000+, the Galaxy Kirin V10 system, the FFTW calculation library version is 3.3.8, the OpenBLAS calculation library version is 3.20.1, local test data is used as the input source, the number of radar beams is 1, the number of pulse accumulations is 1024, the number of sampling points of the echo signal is 4000, and 8 times interpolation is used.

[0086] 2. Experimental content

[0087] 2a) The signal envelope before range alignment, as shown in Figure 2 .

[0088] 2b) The signal envelope after the range alignment method of the present application, as shown in Figure 3 .

[0089] 2c) Envelope cross-correlation processing is performed in parallel using 16 threads, the processing time of 1000 times of operation before and after optimization is counted, and the average result is shown in Table 1:

[0090] Table 1 Comparison of processing time before and after cross-correlation operation optimization

[0091] Pre-optimization envelope cross-correlation processing time (ms) Post-optimization envelope cross-correlation processing time (ms) 10.271 0.232

[0092] 2d) Signal compensation processing is performed in parallel using 16 threads, the processing time of 1000 times of operation before and after optimization is counted, and the average result is shown in Table 2:

[0093] Table 2 Comparison of processing time before and after compensation calculation optimization

[0094] Pre-optimization compensation calculation processing time (ms) Post-optimization compensation calculation processing time (ms) 1.225 0.0274

[0095] 2e) Range alignment processing is performed in parallel using 16 threads, the processing time of 1000 times of operation before and after optimization is counted, and the average result is shown in Table 3:

[0096] Table 3 Comparison of processing time before and after range alignment module optimization

[0097] Pre-optimization processing time (ms) Post-optimization processing time (ms) 909.937 190.296

[0098] 3. Result analysis:

[0099] From Figure 2 It can be seen that the envelopes between the pulses are not aligned;

[0100] From Figure 3 It can be seen that the envelopes between the pulses have been aligned;

[0101] As can be seen from Table 1, under the same hardware platform, the optimized envelope cross-correlation processing performance is improved by about 17.9 times;

[0102] As can be seen from Table 2, under the same hardware platform, the optimized compensation processing performance is improved by about 44.3 times;

[0103] As can be seen from Table 3, through the optimization of cross-correlation and compensation calculation, the whole envelope alignment performance is improved by about 4.7 times, which greatly improves the processing performance.

[0104] The above is only a specific embodiment of the present application, but the protection scope of the present application is not limited to this. Any skilled person in the art can easily think of various equivalent modifications or replacements within the technical range disclosed by the present application, and these modifications or replacements should be covered within the protection scope of the present application.

Claims

1. An envelope correlation based radar signal range alignment implementation method, characterized in that, The method comprises: calculating an original signal envelope of each pulse under the same pulse group; selecting a first pulse as a reference pulse, and calculating an envelope cross-correlation between the envelope of the remaining pulses and the envelope of the reference pulse; the calculating of the envelope cross-correlation between the envelope of the remaining pulses and the envelope of the reference pulse comprises: to a reference pulse envelope do point fourier transform ; to the first remaining pulse envelope do point fourier transform ; On With Inverse Fourier transform of the result of the conjugate vector dot product operation of the envelope cross-correlation result ; obtaining a peak position of the envelope cross-correlation of the original signal; specifically, By traversing the first Envelope cross-correlation of the remaining pulses and the reference pulse Obtaining the peak position thereof ; The original signal is subjected to times interpolation, and the envelope of the interpolated signal is calculated; Obtain the peak position of the envelope cross-correlation of the original signal and its corresponding position after interpolation, and then, using this position as the center, find the positions around it. The envelope cross-correlation is calculated within the range of each sampling point; specifically: By the peak position of the envelope cross-correlation corresponding position after interpolation , ;​ With center, left and right nearby envelope cross-correlation results in the range of sampling points obtained by the following formula wherein, is the interpolated envelope of the reference pulse, is the interpolated envelope of the first pulse, is the interpolated envelope of the second pulse. obtaining a peak position of the envelope cross-correlation of the interpolated signal; compensating the interpolated signal according to the peak position of the envelope cross-correlation of the interpolated signal; specifically, By envelope cross-correlation of the peak position The first interpolated pulse signal compensated frequency domain result may be obtained by the following formula wherein, is the frequency domain result of the first interpolated pulse signal before compensation, is the frequency domain result of the first interpolated pulse signal before compensation, ; The compensated signal is subjected to Decimation.

2. The method of claim 1, wherein the method is implemented by a radar signal distance alignment based on envelope correlation. The original signal is processed times, and the envelope of the interpolated signal is calculated, specifically: performing a Fourier transform on the original signal to obtain ; from the successive insertion of 0s ; ;​​ right The signal envelope is obtained by modulo operation. .

3. The method of claim 1, wherein the method further comprises: The compensated signal is subjected to Decimation, specifically: truncation the first and last data are combined , ; right Perform inverse Fourier transform to obtain the distance-aligned signal ;in The number of sampling points. This indicates rounding up to the nearest integer.

4. A computer system, characterized by The method comprises: one or more processors, a computer readable storage medium for storing one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors implement the method of claim 1.

5. A computer readable storage medium characterized by There are computer executable instructions stored, which are used to implement the method of claim 1 when executed.

6. A computer program product, characterised in that There are computer executable instructions stored, which are used to implement the method of claim 1 when executed.

Images