Space zero-fill interpolation 3D-FFT direction finding method for triangularly arranged digital array

Abstract

The invention relates to the field of radar information, in particular to a triangular array digital array space zero-fill interpolation 3D-FFT direction finding method, which comprises the following steps of: firstly, respectively carrying out range direction FFT on each channel of a triangular array phased array to obtain frequencies of other-source radiation signals and amplitude phases of each channel corresponding to each frequency; the method comprises the following steps: constructing a phase matrix, carrying out spatial zero-fill interpolation, carrying out spatial row-direction FFT and column-direction FFT to obtain a signal spatial spectrum, searching a spectrum peak to obtain a spectrum peak position, and carrying out cubic spline interpolation on the spectrum peak to obtain an accurate arrival direction and an accurate elevation angle of the other-source radiation signal.

Description

Technical Field

[0001] The present invention relates to the field of radar information processing. Background Art

[0002] Radar is a key element in determining the outcome of wars, and it is also facing more and more threats. Active interference is a major threat to modern radars, which makes it difficult for radars to detect real threat targets normally. In order to solve the active interference problem faced by radars, it is of great significance to detect and direction-find the interference signal. At the same time, it is also of great value to measure the arrival direction of signals from other external electronic devices when the radar is silent. The existing array direction-finding methods mainly include spatial spectrum estimation methods such as MUSIC and ESPRIT, but the spatial spectrum estimation algorithm has a huge amount of calculation, and it is difficult to meet the real-time requirements for two-dimensional direction-finding.

[0003] Spatial spectrum estimation methods such as MUSIC and ESPRIT all require the construction of a matrix with the same dimensionality as the number of array elements. They are more commonly used in one-dimensional direction finding with a small number of array elements. When two-dimensional direction finding is required, if the number of array elements is too small, the signal measurement accuracy is insufficient. If the number of array elements is large, a large-dimensional matrix needs to be constructed and eigendecomposition operations need to be performed, which requires a large amount of calculation, is difficult to implement in engineering, and causes processing delays. Summary of the invention

[0004] In view of the problem that a triangular digital array cannot directly perform spatial row and column FFT operations, the present invention proposes a spatial zero-padding interpolation 3D-FFT direction finding method for a triangular digital array. The spatial zero-padding interpolation method is introduced to construct a spatial matrix. Only a small amount of FFT operations and spatial spectrum peak searches are required, which is easy to implement and has strong real-time performance.

[0005] In order to achieve the above technical objectives, the present invention proposes a spatial zero-filling interpolation 3D-FFT direction finding method for a triangular array digital array, and the technical solution includes:

[0006] S1: The radar collects radiation signal data from other sources in the hidden process, and collects data at K points of M×N channels. The first row of channel data is:

[0007] X 11 =[X 11,1 ,X 11,2 ,…X 1,K ], X 12 =[X 12,1 ,X 12,2 ,…X 12,K ], …, X 1N =[X 1N,1 ,X 1N,2 ,…X 1N,K ];

[0008] The channel data of the mth row and nth column is: Xmn=[Xmn,1,Xmn,2,…Xmn, K ], where 1≤m≤M, 1≤n≤N;

[0009] S2: After receiving the M×N channel radiation signal acquisition data, arrange them according to the corresponding spatial position of the channel to obtain M×N×K complex matrix data, perform K-point FFT operation in the distance direction on the M×N channel signal acquisition data, and obtain M×N complex spectrum vectors;

[0010] S3: Select part or all of the M×N complex spectrum vectors and add them up to obtain a real spectrum vector, perform a one-dimensional spectrum peak search on the real spectrum vector, and obtain P spectrum peak positions k1, k2, …k P , then the signal frequency corresponding to ki is

[0011] fi=fc+(ki-K / 2) / K*B ki>K / 2

[0012] fi=fc+ki / K*B ki<K / 2

[0013] Where fc is the carrier frequency, B is the sampling bandwidth, 0≤P≤K, 0≤i≤P;

[0014] S4: Take out P peak positions k1, k2, ...k in sequence P The corresponding complex numbers in the M×N complex spectrum vectors obtain P M×N dimensional complex matrices;

[0015] S5: performing zero-filling interpolation and row-column-direction FFT at each X point on the P M×N dimensional complex matrices to obtain an X×X dimensional complex matrix, where X≥M and X≥2N;

[0016] S6: Perform two-dimensional spectrum peak search on each of the P X×X dimensional complex matrices to obtain the spatial spectrum peak position corresponding to the signal, that is, the spatial angular frequency;

[0017] S7: For the obtained two-dimensional spectrum peak, take values ​​near it and perform cubic spline interpolation, and then search for the maximum position after cubic spline interpolation to obtain the accurate spatial angular frequency. The azimuth direction is The elevation angle is sinθ, where is the azimuth, θ is the elevation angle;

[0018] S8: Calculate the azimuth and elevation of the radiation source according to the spatial angular frequency.

[0019] Different from the traditional array direction finding algorithm, the present invention performs spatial zero-filling interpolation on the triangular array phased array, performs 3D-FFT (three-dimensional FFT) in the range, row and column directions and a cubic spline interpolation to obtain the signal arrival direction. Its significant advantages are:

[0020] 1. Only FFT operation is performed, with small amount of calculation, small processing delay, high real-time performance, and the calculation time of large-scale arrays is reduced by more than 97% compared with the spatial spectrum estimation method;

[0021] 2. Suitable for triangular digital arrays, fast and accurate direction finding, obtaining the frequency, azimuth and elevation of the radiated signal from other sources, with a measurement accuracy of about 1 / 10 of the array wave width.

[0022] The present invention is further described in detail below in conjunction with the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Attached Figure 1 This is a schematic diagram of the zero-filling interpolation of the triangular array of the present invention.

[0024] Attached Figure 2 Principle diagram of 3D-FFT direction finding of triangular array digital array of the present invention DETAILED DESCRIPTION

[0025] The present invention proposes a spatial zero-filling interpolation 3D-FFT direction finding method for a triangular array digital array, which uses three-dimensional FFT calculations in the range direction, spatial row direction, and spatial column direction, and obtains a spatial signal matrix through spatial zero-filling interpolation, and then obtains the final signal arrival direction through spectrum peak search and cubic spline interpolation. The key technology of the present invention lies in: 3D-FFT calculation of the zero-filling interpolation phase matrix.

[0026] The present invention is applicable to radar array direction finding. It performs spatial zero-filling interpolation processing on a triangular array digital array, obtains a rectangular array after zero-filling, and then performs 3D-FFT on the rectangular array after spatial zero-filling, and then performs spectrum peak search and cubic spline interpolation to obtain the signal azimuth and elevation. When the radar is working, a processor chip is used to complete the 3D-FFT direction finding of the triangular array phased array. After the range FFT of the M×N array (M rows and N columns), the signal frequency phase is extracted, and then the spatial zero-filling interpolation is performed to obtain the M×2N phase matrix; the M×2N phase matrix is ​​sequentially subjected to row FFT and column FFT, spectrum peak search and cubic spline interpolation. The spatial zero-filling interpolation of the triangular array is as follows: Figure 1 The basic principle of 3D-FFT direction finding of triangular array digital array is shown in the attached figure. Figure 2 As shown, the specific implementation method is as follows:

[0027] S1: Radar collects radiation signal data from other sources in the dark, and collects data at K points of M×N channels. The first row of channel data is: X 11 =[X 11,1 ,X 11,2 ,…X 1,K ], X 12 =[X 12,1 ,X 12,2 ,…X 12,K ], …, X1N =[X 1N,1 ,X 1N,2 ,…X 1N,K ];

[0028] The channel data of the mth row and nth column (1≤m≤M, 1≤n≤N) is: Xmn=[Xmn,1,Xmn,2,…Xmn, K ];

[0029] S2: After receiving the M×N channel signal acquisition data, the processor arranges them according to the corresponding spatial positions of the channels to obtain M×N×K complex matrix data, and performs K-point FFT operations in the distance direction on the M×N channel signal acquisition data to obtain M×N complex spectrum vectors;

[0030] S3: Select part or all of the M×N complex spectrum vectors and add them up to get a real spectrum vector. Perform a one-dimensional spectrum peak search on the real spectrum vector to get P (0≤P≤K) spectrum peak positions k1, k2, …k P , then the signal frequency corresponding to ki(0≤i≤P) is

[0031] fi=fc+(ki-K / 2) / K*B ki>K / 2

[0032] fi=fc+ki / K*B ki<K / 2

[0033] Where fc is the carrier frequency and B is the sampling bandwidth;

[0034] S4: sequentially extract the ki-th (0≤i≤P) complex number of the corresponding M×N complex spectrum vectors for P (0≤P≤K) frequencies to obtain P M×N dimensional complex matrices;

[0035] S5: Basis Figure 1 In the spatial interpolation zero-filling method of the medium triangular array, zero-filling interpolation is performed on P M×N-dimensional complex matrices respectively to obtain an M×2N-dimensional complex matrix. The values ​​in the M×2N-dimensional complex matrix corresponding to the real array elements are the complex values ​​extracted from the corresponding frequency positions after the range-direction FFT, and the values ​​in the M×2N-dimensional complex matrix corresponding to the virtual array elements are 0;

[0036] S6: Perform row-wise FFT and column-wise FFT on P M×2N dimensional complex matrices respectively:

[0037] ① Perform M X-point complex FFT in the row direction on the M×2N dimensional complex matrix, setting X≥M and X≥2N, to obtain an M×X dimensional complex matrix;

[0038] ② Perform X X-point complex FFT in the column direction on the M×X dimensional complex matrix, setting X ≥ M and X ≥ 2N, to obtain P X×X dimensional complex matrices;

[0039] Where X≥M and X≥2N;

[0040] S7: Perform two-dimensional spectrum peak searches on the P X×X dimensional complex matrices respectively to obtain the spatial spectrum peak position corresponding to the signal, that is, the spatial angular frequency;

[0041] S8: For the obtained two-dimensional spectrum peak, take values ​​near it and perform cubic spline interpolation, and then search for the maximum position after cubic spline interpolation to obtain the accurate spatial angular frequency. The azimuth direction is The elevation angle is sinθ, where is the azimuth and θ is the elevation angle.

[0042] S9: The azimuth and elevation of the radiation source can be calculated based on the spatial angular frequency.

[0043] Taking a 32×32 triangular digital array as an example, there are 1024 channels in total, and each channel is sampled at 64 points. Then, 1024 64-point FFT operations are performed in the distance direction. Assuming that there are two external radiation signals, two 32×32 phase matrices are constructed; two 32×64 phase matrices are obtained after spatial zero padding; each matrix can be subjected to 128-point FFT operations in the row and column directions respectively, so a total of 320 128-point FFT operations are performed to obtain two 128×128 signal spatial spectrum matrices; then the two signal spatial spectra are searched for two-dimensional spectrum peaks to obtain the spatial angular frequency corresponding to the signal; for the obtained two-dimensional spectrum peaks, values ​​are taken near them for cubic spline interpolation, and the maximum value position is searched after cubic spline interpolation to obtain the accurate spatial angular frequency. A total of 3D-FFTs are performed in the distance, row, and column directions, including 1024 64-point FFTs and 320 128-point FFTs. The implementation is easy and the calculation amount is small. The total processing time is measured to be less than 15ms, and the angle direction finding error is less than 0.3°.

Claims

1. A spatial zero-filling interpolation 3D-FFT direction finding method for a triangular array digital array, characterized in that: S1: The radar collects radiation signal data from other sources in the hidden process, and collects data at K points of M×N channels. The first row of channel data is: X 11 =[X 11,1 ,X 11,2 ,…X 1,K ],X 12 =[X 12,1 ,X 12,2 ,…X 12,K ],…,X 1N =[X 1N,1 ,X 1N,2 ,…X 1N,K ]; The channel data of row m and column n is: mn =[X mn,1 ,X mn,2 ,…X mn,K ], where 1≤m≤M, 1≤n≤N; S2: After receiving the M×N channel radiation signal acquisition data, arrange them according to the corresponding spatial position of the channel to obtain M×N×K complex matrix data, perform K-point FFT operation in the distance direction on the M×N channel signal acquisition data, and obtain M×N complex spectrum vectors; S3: Select part or all of the M×N complex spectrum vectors and add them up to obtain a real spectrum vector, perform a one-dimensional spectrum peak search on the real spectrum vector, and obtain P spectrum peak positions k1, k2, …k P , then k i The corresponding signal frequency is: f i =f c +(k i -K / 2) / K*B k i >K / 2 f i =f c +k i / K*B k i <K / 2 where f c is the carrier frequency, B is the sampling bandwidth, 0≤P≤K, 0≤i≤P; S4: Take out P peak positions k1, k2, ...k in sequence P The corresponding complex numbers in the M×N complex spectrum vectors obtain P M×N dimensional complex matrices; S5: performing zero-filling interpolation and row-column-direction FFT at each X point on the P M×N dimensional complex matrices to obtain an X×X dimensional complex matrix, where X≥M and X≥2N; S6: Perform two-dimensional spectrum peak search on each of the P X×X dimensional complex matrices to obtain the spatial spectrum peak position corresponding to the signal, that is, the spatial angular frequency; S7: For the obtained two-dimensional spectrum peak, take values ​​near it and perform cubic spline interpolation, and then search for the maximum position after cubic spline interpolation to obtain the accurate spatial angular frequency. The azimuth direction is The elevation angle is sinθ, where is the azimuth, θ is the elevation angle; S8: Calculate the azimuth and elevation of the radiation source according to the spatial angular frequency.

2. The spatial zero-filling interpolation 3D-FFT direction finding method of a triangular array digital array according to claim 1, characterized in that: S1: P M×N-dimensional complex matrices are interpolated with zeros to obtain an M×2N-dimensional complex matrix. The values ​​in the M×2N-dimensional complex matrix corresponding to the real array elements are the complex values ​​extracted from the corresponding spectrum peak positions after the range FFT. The values ​​in the M×2N-dimensional complex matrix corresponding to the virtual array elements are 0. S2: Perform M X-point complex FFTs in the row direction on the P M×2N-dimensional complex matrices respectively, setting X≥M and X≥2N to obtain M×X-dimensional complex matrices; S3: Perform X X-point complex FFTs in the column direction on the P M×X dimensional complex matrices respectively, setting X≥M and X≥2N, to obtain P X×X dimensional complex matrices.

Images