A low computational complexity adaptive beamforming method for dual-polarization arrays

Abstract

This invention discloses a low-computational-complexity adaptive beamforming method for dual-polarization arrays. The steps include: constructing a steering vector and a beam conversion matrix; whitening preprocessing and beam domain conversion; reconstructing the received signal and steering vector in the beam domain; adaptive beamforming in the beam domain; and calculating and comparing the output SINR. This invention solves the problems of high computational load and complexity in existing element-domain adaptive beamforming methods. While ensuring good anti-interference performance, it reduces the dimension of the inverse matrix required in adaptive beamforming, thereby reducing the computational load and system complexity. It achieves better anti-interference performance under low snapshot conditions, with faster convergence speed, and more closely meets the stringent real-time and efficiency requirements of radar systems in practical applications.

Description

Technical Field

[0001] This invention belongs to the field of radar technology, and more specifically relates to a low-computational-complexity dual-polarized array adaptive beamforming (ADBF) method in the field of radar anti-jamming technology. This invention can be used in complex and variable jamming environments, employing weighted spatial filtering of antenna array elements to suppress jamming signals that differ from the target signal. Background Technology

[0002] Adaptive beamforming is a key technology in radar signal processing. It uses weighted antenna elements for spatial filtering to suppress interference signals different from the target signal, with the adaptive weight vector updated dynamically in real time. However, current dual-polarization array adaptive beamforming methods have limitations. First, existing methods mostly process in the element domain. While dimensionality reduction techniques can reduce computation, they are highly demanding on the number of snapshots, achieving ideal performance only with a large number of snapshots. Second, in scenarios with a low number of snapshots, this method suffers from two major drawbacks: computational load is difficult to control effectively, increasing the computational burden on the radar system and affecting operational efficiency; anti-interference performance converges slowly, making it difficult to quickly adapt to complex interference environments, failing to guarantee stable and efficient radar system operation, and not meeting practical application requirements. Third, in radar search mode, target polarization information cannot be known in advance and is time-varying. The target polarization information preset when calculating the weights in existing dual-polarization array adaptive beamforming methods may deviate significantly from the actual polarization.

[0003] In their paper "A Time-Division Dual-Polarization Array ADBF Processing Method" (Journal of Aerospace Early Warning Research, Vol. 38, No. 4, August 2024), Liu Hao et al. proposed an adaptive beamforming method combining polarization and spatial domains. This method innovatively designs an alternating dual-polarization array, dividing it into multiple polarization subarray regions. First, using different subarray transformation matrices, the received signals and steering vectors under horizontal and vertical polarization are derived after dimensionality reduction. Then, adaptive beamforming and pulse compression processing are performed. Finally, the output signal-to-interference-plus-noise ratio (SINR) of the two polarization methods is compared, and the data with the larger value is selected as the final output. This method provides a new approach to anti-jamming for dual-polarization radar, but it has shortcomings. Its dimensionality reduction in the array element domain is highly dependent on the number of snapshots; only with a large number of snapshots can the output SINR be significantly improved, resulting in ideal anti-jamming performance. In scenarios where the number of snapshots is limited, the anti-jamming effect is unsatisfactory.

[0004] Xi'an University of Electronic Science and Technology disclosed a method for joint anti-jamming in the polarization-spatial domain for dual-polarized radar in its patent application "A Joint Anti-jamming Method for Polarization-Spatial Domain for Dual-Polarized Radar" (Application No.: 202311811870.2, Application Date: 2023.12.26, Publication No.: CN 117826088 A). The implementation steps are as follows: estimating the polarization-spatial domain covariance matrix based on the received signals from the horizontal and vertical polarization arrays of the dual-polarized radar; constructing a desired signal steering vector corresponding to the transmitting polarization mode and transmission angle of the dual-polarized radar; calculating the polarization-spatial domain adaptive weight vector based on the polarization-spatial domain covariance matrix and the desired signal steering vector; and using the polarization-spatial domain adaptive weight vector to perform a weighted summation of the received signals to obtain the joint anti-jamming output in the polarization-spatial domain. This method overcomes the shortcomings of existing technologies in setting the desired signal polarization mode when the target polarization mode is unknown, and addresses the incompatibility with multiple targets caused by unreasonable setting of the desired signal polarization mode in existing technologies, thus improving the anti-jamming performance of dual-polarized radar. However, this method falls under the category of array element domain processing and has the drawback of array element domain adaptive beamforming. Summary of the Invention

[0005] The purpose of this invention is to address the shortcomings of the prior art by proposing a low-computational-complexity adaptive beamforming method for dual-polarized arrays. This method aims to solve the problems of existing technologies that only perform anti-interference processing in the array element domain, which has extremely stringent requirements on the number of snapshots and consumes a large amount of additional computing resources and time.

[0006] To achieve the above objectives, the technical approach of this invention is as follows: This invention converts the signal from the element space to the beam space through beam domain conversion, transforming the received signal of the dual-polarized array from the element domain to the beam domain, thereby reducing the dimension of the inverse matrix. This operation has two major advantages: First, compared to adaptive beamforming in the element domain, adaptive beamforming in the beam domain can achieve better anti-interference performance under low snapshot conditions, solving the problem of the dependence of adaptive beamforming in the element domain on the number of snapshots; Second, it reduces the dimension of the inverse matrix, thereby reducing the computational load and system complexity of the algorithm, and accelerating the convergence speed of anti-interference performance, which is closer to the stringent requirements of real-time performance and efficiency for radar systems in actual application scenarios. This invention performs channel-specific processing on the received signal of the dual-polarized array, using the output of the channel with the higher SINR as the final result of the dual-polarized array signal processing. The higher the SINR of the channel output, the better the matching degree between the channel's polarization mode and the target's polarization parameters. Therefore, the channel-specific processing method does not need to estimate the target's polarization parameters in real time. Accordingly, it solves the problem of polarization information deviation caused by the pre-set target polarization information in the process of calculating weights in existing dual-polarization array adaptive beamforming.

[0007] To achieve the above objectives, the technical solution adopted by the present invention includes the following steps:

[0008] Step 1: Construct horizontal and vertical polarization arrays to receive signals;

[0009] Step 2, construct the horizontal and vertical polarization steering vectors;

[0010] Step 3: Construct the beam conversion matrices for the horizontal polarization array and the vertical polarization array, respectively;

[0011] Step 4: By whitening preprocessing the beam conversion matrix, the array received signals and steering vectors of the corresponding array elements in the horizontal and vertical polarization domains are converted to the beam domain respectively.

[0012] Step 5: Reconstruct the beam domain received signal and the beam domain steering vector;

[0013] Step 6: Perform different adaptive weighting processes on the received signals of the beam domain dual-polarization array to obtain horizontal and vertical polarization output data.

[0014] Step 7: Compare the output SINR of the horizontal and vertical polarization output data, and take the channel with the larger output SINR as the final result of the dual-polarization array signal processing.

[0015] Furthermore, the steps for constructing the horizontally polarized array and the vertically polarized array to receive signals are as follows:

[0016] Based on the array signal processing model, construct the horizontally polarized array to receive the signal X. H X H The signal is received by an N1×L dimensional horizontal polarization array, where N1 is equal to the number of elements in the horizontal polarization array and L is equal to the total number of snapshots.

[0017] Based on the array signal processing model, construct the vertically polarized array receiving signal X. V ,X V The signal is received by an N2×L dimensional vertical polarization array, where the value of N2 is equal to the number of elements in the vertical polarization array.

[0018] Furthermore, the horizontal and vertical polarization steering vectors are obtained through the beam pointing angle of the dual-polarization array:

[0019] When the beam pointing angle of the dual-polarized array is θ1, the N1×1 dimension horizontal polarized array steering vector a is obtained. H (θ1) and N2×1 dimensional vertical polarization array steering vector a V (θ1).

[0020] Furthermore, the beam conversion matrix is ​​obtained by the following steps:

[0021] The angular range Θ of all radar sources is uniformly divided into M directions with angles σ1,...,σ1. M ;

[0022] The M directional guide vectors of the horizontal polarization array are combined into the beam conversion matrix of the horizontal polarization array: T1 = [a H (σ1),a H (σ2),...,a H (σ M )], T1 is an N1×M dimensional matrix;

[0023] The M directional guide vectors of the vertical polarization array are combined into the beam conversion matrix of the vertical polarization array: T2 = [a V (σ1),a V (σ2),...,a V (σ M )], T2 is an N2×M dimensional matrix.

[0024] Furthermore, the steps of the whitening pretreatment are as follows:

[0025] The beam conversion matrix T1 of the horizontal polarization array is preprocessed by whitening to obtain the preprocessed beam conversion matrix. T3 is an N1×M dimensional matrix, [·] H This represents the matrix conjugate transpose operation;

[0026] The beam conversion matrix T2 of the vertical polarization array is preprocessed with whitening to obtain the preprocessed beam conversion matrix. T4 is an N2×M dimensional matrix.

[0027] Furthermore, the step of converting the array received signal and steering vector in the horizontally polarized and vertically polarized array element domains to the beam domain is as follows:

[0028] Receive signal X from the horizontal polarization array H The signal is converted to the beam domain to obtain the horizontally polarized array received signal in the beam domain. X′ H Let L represent an M×L dimensional matrix, where the value of L is equal to the total number of snapshots.

[0029] The horizontal polarization steering vector a H (θ1) is transformed to the beam domain to obtain the beam domain horizontal polarization steering vector. a′ H (θ1) represents an M×1 dimensional vector;

[0030] The vertical polarization array receives signal X V The signal is converted to the beam domain to obtain the beam domain vertical polarization array received signal. X′V Represents an M×L dimensional matrix;

[0031] The vertical polarization guide vector a V (θ1) is transformed to the beam domain to obtain the beam domain vertical polarization steering vector. a′ V (θ1) represents an M×1 dimensional vector.

[0032] Furthermore, the steps for recombining the beam domain received signal and the beam domain steering vector are as follows:

[0033] The horizontal polarization array in the beam domain receives the signal X′. H and vertical polarization array receive signal X′ V The dual-polarized array can be combined into a beam domain to receive the signal X′:

[0034]

[0035] Where X′ represents a 2M×L dimensional matrix, where 2M is less than N, and N is equal to the sum of the number of horizontal polarization array elements and the number of vertical polarization array elements.

[0036] The beam domain horizontal polarization steering vector a′ H (θ1) and the zero vector are combined to obtain the beam domain horizontal polarization hybrid steering vector a. MH (θ1):

[0037]

[0038] in, Let a be an N3×1 dimensional zero vector, where N3 is equal to M. MH (θ1) represents a 2M×1 guiding vector;

[0039] The zero vector and the beam domain vertical polarization guide vector a′ V (θ1) is combined to obtain the beam domain vertical polarization hybrid steering vector a. MV (θ1):

[0040]

[0041] in, This represents the zero vector of dimension N4×1, where N4 is equal to M,a MV (θ1) represents a 2M×1 guiding vector.

[0042] Furthermore, the covariance matrix is:

[0043]

[0044] in, Let C represent the 2M×2M beam domain polarization domain-spatial domain joint covariance matrix, and let C represent the 2M×l beam domain dual-polarization array interference snapshot selection matrix. Let l represent the number of columns in matrix C. Its value represents the number of data extracted from the total number L of snapshots of the dual-polarization array receiving signal X′ in the beam domain. Its value range is [2M,L]. The sequence number of the selected snapshot is not equal to the sequence number of the snapshot where the target is located.

[0045] Furthermore, the adaptive weighting process includes the following steps:

[0046] Using the joint covariance matrix of polarization domain and spatial domain and the horizontal polarization hybrid steering vector a in the beam domain MH (θ1) Calculate the horizontal polarization adaptive weight vector in the beam domain. w H The horizontal polarization output data is obtained from a 2M×1 dimensional vector. Then, take the square of the modulus to obtain the instantaneous power P of the signal. H ;

[0047] Using the joint covariance matrix of polarization domain and spatial domain and the vertical polarization hybrid steering vector a in the beam domain MV (θ1), calculate the vertical polarization adaptive weight vector in the beam domain. w V The vertical polarization output data is obtained from a 2M×1 dimensional vector. Then, take the square of the modulus to obtain the instantaneous power P of the signal. V .

[0048] Furthermore, the output SINR of the horizontal and vertical polarization output data is obtained by the following steps:

[0049] Based on the instantaneous power P of the horizontal polarization output signal H Calculate the output SINR, the output SINR of the horizontal polarization output data. H In decibel form:

[0050]

[0051] Among them, P HM P represents the power of the horizontally polarized output target signal. HIN This represents the sum of interference power and noise power in the horizontally polarized output data;

[0052] Based on the instantaneous power P of the vertical polarization output signal V Calculate the output SINR, the output SINR of the vertical polarization output data. V In decibel form:

[0053]

[0054] Among them, P VM P represents the power of the vertically polarized output target signal. VIN This represents the sum of interference power and noise power in the vertically polarized output data.

[0055] Compared with the prior art, the present invention has the following advantages:

[0056] First, this invention transforms the signal from the element space to the beam space through beam domain conversion, overcoming the shortcomings of the element domain adaptive beamforming method in terms of large computational load and high complexity. This invention enables the transformation from the element domain to the beam domain, while ensuring good anti-interference performance of the method, reducing the dimension of the inverse matrix required in adaptive beamforming, reducing the computational load and system complexity of the algorithm, and facilitating engineering implementation.

[0057] Second, the present invention performs adaptive beamforming in the beam domain. Due to the reduction in the dimension of the inverse matrix, it overcomes the dependence of the adaptive beamforming method in the element domain on the number of snapshots. This allows the method of the present invention to achieve better anti-interference performance under low snapshot conditions, with faster convergence speed, and is closer to the stringent requirements of real-time performance and efficiency of radar systems in actual application scenarios.

[0058] Third, the present invention performs adaptive beamforming in different channels and compares the corresponding anti-interference performance. The anti-interference performance measures the degree of fit between the polarization mode of the channel and the polarization state of the target. This effectively solves the problem of polarization information deviation caused by pre-setting the target polarization information when calculating weights in the existing dual-polarization array adaptive beamforming technology. This makes the present invention more flexible and accurate in adapting to the polarization characteristics of different targets in practical applications. Attached Figure Description

[0059] Figure 1 is a flow chart of the present invention;

[0060] Figure 2 This is a simulation diagram of the present invention, wherein, Figure 2 (a) is a comparison of the adaptive beamforming output results of the array element domain and beam domain under low-speed simulation experiment 1 of the present invention. Figure 2 (b) is a comparison diagram of the adaptive beamforming output SINR of two polarization modes in the element domain horizontal polarization and beam domain under different snapshot numbers in the simulation experiment 2 of this invention. Figure 2 (c) is a graph showing the change of SINR output with the interference angle in the simulation experiment 3 of the present invention and the prior art method. Detailed Implementation

[0061] The present invention will be further described in detail below with reference to the accompanying drawings and embodiments.

[0062] Reference Figure 1 The specific implementation steps of the embodiments of the present invention will be described in further detail below.

[0063] Step 1: Construct an array to receive signals.

[0064] Step 1.1: Based on the array signal processing model, obtain the horizontally polarized array received signal X. H :

[0065]

[0066] Where there is one target signal, K represents the number of interfering signals, and s t Let s represent the complex envelope vector of the target. t =[ζ t1 ,ζ t2 ,...,ζ tr ,...,ζ tL ],1≤r≤L,ζ tr The complex envelope of the target is represented by the value s, which indicates that the target exists in the r-th snapshot. t The remaining L-1 elements are all 0, indicating that s ik The complex envelope vector of the k-th interference echo, k = 1, 2, ..., K; X H This represents the received signal of an N1×L dimensional horizontally polarized array, where N1 represents the number of elements in the horizontally polarized array, L represents the total number of snapshots, and a H (θ t ) represents a horizontally polarized array pointing in the direction θ. t The guide vector, θ represents the relative complex envelope of the k-th interference echo received by the horizontally polarized array. ik Indicates the direction of the k-th interference, a H (θ ik ) represents a horizontally polarized array pointing in the direction θ. ik The guiding vector, n H It is represented as the noise matrix of an N1×L dimensional horizontally polarized array.

[0067] Step 1.2: Based on the array signal processing model, construct the vertical polarization array received signal X. V :

[0068]

[0069] X V This represents the signal received by an N2×L dimensional vertically polarized array, where N2 represents the number of elements in the vertically polarized array, and a V (θ t ) represents the vertical polarization array pointing in the direction θ. tThe guide vector, Let a represent the relative complex envelope of the k-th interference echo of the vertically polarized array. V (θ ik ) represents the vertical polarization array pointing in the direction θ. ik The guiding vector, n V It is represented as the noise matrix of an N2×L dimensional vertical polarization array.

[0070] Since the geometry of the array determines the characteristics of the steering vector, horizontally polarized and vertically polarized array antennas can be arranged in any configuration. In this invention, for ease of explanation, we use an equidistant linear array as an example to illustrate the arrangement of horizontally and vertically polarized array antennas. It should be noted that this example of an equidistant linear array is universal; that is, its principles and methods can be extended to other horizontally and vertically polarized array antennas with arbitrary configurations.

[0071] Step 2, construct the guide vector.

[0072] When the beam pointing angle of the dual-polarized array is θ1, the N1×1 dimension horizontal polarized array steering vector a is obtained. H (θ1) and N2×1 dimensional vertical polarization array steering vector a V (θ1).

[0073] Step 3: Construct the beam conversion matrix.

[0074] Step 3.1: Since radar can obtain the approximate angular range Θ of all signal sources in actual engineering, Θ is uniformly divided into M directions with angles σ1,...,σ1. M .

[0075] Step 3.2: Using the angles of M uniformly divided directions, generate the guiding vector a of the horizontal polarization array in M ​​directions. H (σ i ), i=1,...,M, combine the M directional steering vectors into a beam conversion matrix T1=[a H (σ1),a H (σ2),...,a H (σ M )], T1 is an N1×M dimensional matrix.

[0076] Step 3.3: Using the angles of M uniformly divided directions, generate the guiding vectors a of the vertical polarization array in M ​​directions. V (σ i ), i=1,...,M, combine the M directional steering vectors into a beam conversion matrix T2=[a V (σ1),a V (σ2),...,aV (σ M )], T2 is an N2×M dimensional matrix.

[0077] Step 4: Whitening preprocessing and beam domain conversion.

[0078] Step 4.1: Perform whitening preprocessing on the beam conversion matrix T1 of the horizontal polarization array from Step 3.2 to obtain the whitening preprocessed beam conversion matrix. T3 is an N1×M dimensional matrix, [·] H This represents the matrix conjugate transpose operation.

[0079] Step 4.2: Perform whitening preprocessing on the beam conversion matrix T2 of the vertical polarization array from Step 3.3 to obtain the whitening preprocessed beam conversion matrix. T4 is an N2×M dimensional matrix.

[0080] Step 4.3, receive the horizontally polarized array signal X from step 1.1. H The signal is converted to the beam domain to obtain the horizontally polarized array received signal in the beam domain. X′ H This represents an M×L dimensional matrix.

[0081] Step 4.4, adjust the horizontal polarization steering vector a from step 2. H (θ1) is transformed to the beam domain to obtain the beam domain horizontal polarization steering vector. a′ H (θ1) represents an M×1 dimensional vector.

[0082] Step 4.5, receive the vertical polarization array signal X from step 1.2. V The signal is converted to the beam domain to obtain the beam domain vertical polarization array received signal. X′ V This represents an M×L dimensional matrix.

[0083] Step 4.6, adjust the vertical polarization steering vector a from step 2. V (θ1) is transformed to the beam domain to obtain the beam domain vertical polarization steering vector. a′ V (θ1) represents an M×1 dimensional vector.

[0084] Step 5: Reconstruct the beam domain array received signal and the beam domain steering vector.

[0085] Step 5.1, receive the horizontally polarized array signal X′ from the beam domain in step 4.3. H And the vertical polarization array received signal X′ in step 4.5 V The dual-polarized array can be combined into a beam domain to receive the signal X′:

[0086]

[0087] Where X′ represents a 2M×L dimensional matrix, where 2M is less than N, and N is equal to the sum of the number of horizontally polarized array elements and the number of vertically polarized array elements.

[0088] Step 5.2, adjust the beam domain horizontal polarization steering vector a′ from step 4.4. H (θ1) and the zero vector are combined to obtain the beam domain horizontal polarization hybrid steering vector a. MH (θ1):

[0089]

[0090] in, Let a be an N3×1 dimensional zero vector, where N3 is equal to M. MH (θ1) represents a 2M×1 guiding vector.

[0091] Step 5.3, combine the zero vector and the beam domain vertical polarization guide vector a′ from step 4.6. V (θ1) is combined to obtain the beam domain vertical polarization hybrid steering vector a. MV (θ1):

[0092]

[0093] in, This represents the zero vector of dimension N4×1, where N4 is equal to M,a MV (θ1) represents a 2M×1 guiding vector.

[0094] Step 6, adaptive beamforming in the beam domain.

[0095] Step 6.1: Estimate the joint covariance matrix of the polarization domain and spatial domain using the received signal X′ from the dual-polarization array in the beam domain as described in Step 5.1.

[0096]

[0097] in, Let C represent the 2M×2M beam domain polarization domain-spatial domain joint covariance matrix, and let C represent the 2M×l beam domain dual-polarization array interference snapshot selection matrix. Let l represent the number of columns in matrix C. Its value represents the number of data extracted from the total number L of snapshots of the dual-polarization array receiving signal X′ in the beam domain. Its value range is [2M,L], and the sequence number of the selected snapshot is not equal to the sequence number of the snapshot where the target is located.

[0098] Step 6.2: Based on the principle of adaptive beamforming, utilize the joint covariance matrix of polarization domain and spatial domain from step 6.1. And step 5.2 horizontal polarization hybrid steering vector a in the beam domain MH (θ1) Calculate the horizontal polarization adaptive weight vector in the beam domain w H It is represented as a 2M×1 dimensional vector.

[0099] Step 6.3: Based on the principle of adaptive beamforming, utilize the joint covariance matrix of polarization domain and spatial domain from step 6.1. And step 5.3 vertical polarization hybrid steering vector a in the beam domain MV (θ1) Calculate the vertical polarization adaptive weight vector in the beam domain. w V It is represented as a 2M×1 dimensional vector.

[0100] Step 6.4, using the horizontal polarization adaptive weight vector w in the beam domain from step 6.2. H Adaptive weighting is performed to obtain horizontally polarized output data. Then, take the square of the modulus to obtain the instantaneous power P of the signal. H .

[0101] Step 6.5, using the vertical polarization adaptive weight vector w in the beam domain from Step 6.3. V Adaptive weighting is performed to obtain vertically polarized output data. Then, take the square of the modulus to obtain the instantaneous power P of the signal. V .

[0102] Step 7: Calculate and compare the output SINR.

[0103] Step 7.1, based on the instantaneous power P of the horizontally polarized output signal in step 6.4 H Calculate the output SINR, the output SINR of the horizontal polarization output data. H In decibel form:

[0104]

[0105] Where lg(·) represents the logarithmic operation to the base 10; P HM This represents the power of the horizontally polarized output target signal, and its value is equal to P. H The maximum value in; P HIN This represents the sum of interference power and noise power in the horizontally polarized output data, and its value is equal to P. H It does not include the average value of the sum of the target signal power values.

[0106] Step 7.2, based on the instantaneous power P of the vertically polarized output signal in step 6.5V Calculate the output SINR, the output SINR of the vertical polarization output data. V In decibel form:

[0107]

[0108] Among them, P VM This represents the power of the vertically polarized output target signal, and its value is equal to P. V The maximum value in; P VIN This represents the sum of interference power and noise power in the vertically polarized output data, and its value is equal to P. V It does not include the average value of the sum of the target signal power values.

[0109] Step 7.3: Compare the output SINR of the horizontal polarization output data in Step 7.1. H The output SINR of the vertical polarization output data in step 7.2 V Take the larger of the data Y. H Or Y V As the final output of the dual-polarization array.

[0110] The effects of this invention will be further illustrated below with simulation experiments:

[0111] 1. Simulation experimental conditions.

[0112] The hardware platform for the simulation experiment of this invention is: Intel(R) Core(TM) i7-12700H CPU with a main frequency of 2.30GHz and 16.00GB of memory.

[0113] The software platform for the simulation experiment of this invention is: Windows 11 operating system and MATLAB R2023a.

[0114] The radar used in the simulation experiment of this invention is a dual-polarization radar, with 13 elements in its horizontal polarization array and 10 elements in its vertical polarization array. In the antenna arrangement, the element spacing of the horizontal and vertical polarization arrays is a half-wavelength equidistant linear array, with the horizontal polarization array located to the left of the vertical polarization array. The distance between the rightmost element of the horizontal polarization array and the leftmost element of the vertical polarization array is half a wavelength.

[0115] 2. Simulation content and result analysis.

[0116] This invention consists of three simulation experiments.

[0117] Simulation Experiment 1 employs the beam domain horizontal polarization of this invention, the beam of this invention, and the method of existing technology to perform anti-jamming processing on the dual-polarized radar received signal under low snapshot conditions. The results obtained after anti-jamming processing are plotted as follows: Figure 2 (a) shows the three curves.

[0118] The simulation experiment 1 of this invention involves a single target in the interference scenario, and the polarization phase descriptor of the target echo signal is (ω s1 ,η s1 The target is located at the 600th snapshot, with an angle of -1° and a signal-to-noise ratio of 25dB. There are two interfering signals; the polarization phase descriptor for interfering signal 1 is (ω...). i1 ,η i1 )=(50,60), the angle of interference signal 1 is 1°, the interference signal 1 has an interference-to-noise ratio of 100dB, and the polarization phase descriptor of interference signal 2 is (ω i2 ,η i2 Given (45,0), the angle of interference signal 1 is -15°, and the interference-to-noise ratio of interference signal 2 is 80dB. Therefore, interference signal 1 is the main lobe interference, and interference signal 2 is the side lobe interference. The angle range Θ of the constructed beam conversion matrix is ​​-20° to 20°, divided into 5 uniform pointing angles. The pointing angle of the dual-polarized array beam is set to 0°. The snapshot index for estimating the covariance matrix is ​​selected between 1000 and 1100, and the snapshot number is 100, i.e., low snapshots are used for estimation.

[0119] Simulation Experiment 2 employs the beam domain horizontal polarization of this invention, the beam of this invention, and the method of existing technologies. Monte Carlo experiments are conducted at different snapshot numbers to perform anti-interference processing on the dual-polarized radar received signal under interference scenarios and calculate the corresponding output SINR, yielding results as follows: Figure 2 (b) shows the output SINR variation curves under different snapshot numbers.

[0120] The simulation conditions in simulation experiment 2 of this invention are basically the same as those in simulation experiment 1, except that the number of snapshots in simulation experiment 2 is not constant. In simulation experiment 2, the snapshot index for estimating the covariance matrix starts from 1000, the value range of the snapshot is [20, 2000], the snapshot interval is 20, and 500 Monte Carlo experiments are performed.

[0121] Simulation Experiment 3 employs the methods of this invention and existing technologies to perform anti-interference processing on radar received signals at different interference angles. The output SINR of both methods is statistically analyzed using Monte Carlo experiments, and the change in output SINR with interference angle is plotted as shown below. Figure 2 The two curves shown in (c)

[0122] In simulation experiment 3 of this invention, the number of snapshots is 100, and the snapshot index for estimating the covariance matrix is ​​selected between 1000 and 1100; the number of targets is 1, and the polarization phase descriptor of the target echo signal is (ω s1 ,η s1 )=(10,0), target direction is 0°, target signal-to-noise ratio is 25dB; number of jammers is 1, jammer polarization phase descriptor is (ω i1 ,η i1 = (50,10), the interference-to-noise ratio is 60dB, the interference direction is set from -20° to 20° and traversed in 1° increments. For each interference step, 500 Monte Carlo tests are performed to statistically analyze the output SINR of existing technologies and this invention.

[0123] The existing technology in the three simulation experiments refers to the method proposed by Liu Hao et al. in their published paper "A Time-Division Dual-Polarization Array ADBF Processing Method" (Journal of Aerospace Early Warning Research, Vol. 38, No. 4, August 2024).

[0124] The effects of the present invention will be further described below with reference to simulation diagrams.

[0125] Figure 2 In (a), the horizontal axis represents the number of snapshots, the vertical axis represents the amplitude, and the unit is dB. The dashed line represents the prior art, the solid line represents the beam domain horizontal polarization method of the present invention, and the dotted line represents the beam domain vertical polarization method of the present invention.

[0126] from Figure 2 As shown in (a), among the curves obtained by the three methods, the target signal amplitude obtained by the anti-interference processing result using the existing technology is the smallest, and its amplitude is smaller than that obtained by the beam domain horizontal polarization method and the beam domain vertical polarization method of the present invention. This indicates that the existing technology has the worst anti-interference effect under low snapshot conditions, while the beam domain horizontal polarization method of the present invention has the best anti-interference effect. Since the target echo polarization parameters are set to be mainly horizontally polarized in the simulation, the beam domain horizontal polarization method of the present invention has a better anti-interference effect than the beam domain vertical polarization method of the present invention. This indicates that the polarization mode of the horizontal polarization processing channel matches the polarization parameters of the target more closely, therefore the final output of the dual-polarization array is the processing result of the horizontal polarization processing channel.

[0127] Figure 2 In (b), the horizontal axis represents the number of snapshots, the vertical axis represents the output SINR, and the unit is dB. The dashed line represents the prior art, the solid line represents the beam domain horizontal polarization method of the present invention, and the dotted line represents the beam domain vertical polarization method of the present invention.

[0128] from Figure 2As can be seen in (b), firstly, compared to existing technologies, the method of this invention requires fewer snapshots to achieve stable output SINR, and the output SINR curve of this invention converges faster. This is because the operation of transforming from the element domain to the beam domain reduces the dimension of the inverse matrix, thereby reducing the computational load and system complexity of the algorithm, and accelerating the convergence speed of anti-interference performance, responding more quickly to changes in interference. Secondly, under low snapshot conditions, the output SINR of the beam domain horizontal polarization method of this invention is greater than that of existing methods, indicating that the anti-interference performance of this invention is superior to existing methods under low snapshot conditions. Therefore, the method of this invention is easier to implement in practical engineering.

[0129] Figure 2 In (c), the horizontal axis represents the interference angle in degrees, and the vertical axis represents the output SINR in dB. Figure 2 In (c), the solid lines represent the prior art, and the dashed lines represent the method of the present invention.

[0130] from Figure 2 As can be seen in (c), under low snapshot conditions, regardless of whether the interference signal is located in the main lobe region or the side lobe region, the output SINR of the existing technology is lower than that of the method of the present invention. Therefore, under low snapshot conditions, the anti-interference performance of the method of the present invention is better than that of the existing technology.

Claims

1. A method for adaptive beamforming of a dual-polarization array with low computational complexity, characterized in that, By converting the received signal and the guide vector beam domain, the subsequent processing flow is transformed from the element domain to the beam domain. A dual-polarized array is used to receive the signal and perform channel-specific processing, with adaptive beamforming performed in both the horizontal and vertical polarization channels. The steps of this beamforming method include the following: Step 1: Construct horizontal and vertical polarization arrays to receive signals; Step 2, construct the horizontal and vertical polarization steering vectors; Step 3: Construct the beam conversion matrices for the horizontal polarization array and the vertical polarization array, respectively; Step 4: By whitening preprocessing the beam conversion matrix, the received signals and steering vectors of the corresponding array element domains of horizontal polarization and vertical polarization are converted to the beam domain respectively. Step 5: Reconstruct the beam domain received signal and the beam domain steering vector; Step 6: Perform different adaptive weighting processes on the received signals of the dual-polarization array in the beam domain to obtain horizontal and vertical polarization output data. Step 7: Compare the output SINR of the horizontal and vertical polarization output data, and take the channel with the larger output SINR as the final result of the dual-polarization array signal processing.

2. The forming method according to claim 1, characterized in that, The steps for constructing the horizontally polarized array and the vertically polarized array to receive signals described in step 1 are as follows: The first step is to construct the horizontally polarized array receiving signal based on the array signal processing model. , for Horizontal polarization received signal, The value of is equal to the number of elements in the horizontally polarized array. The value of is equal to the total number of snapshots; The second step is to construct a vertically polarized array to receive signals based on the array signal processing model. , for Vertical polarization received signal, The value of is equal to the number of vertically polarized array elements.

3. The forming method according to claim 2, characterized in that, The horizontal and vertical polarization steering vectors mentioned in step 2 are obtained through the beam pointing angle of the dual-polarization array: When the beam pointing angle of the dual polarization array is At that time, Horizontal polarization array steering vector and Vertical polarization array guide vector .

4. The forming method according to claim 3, characterized in that, The beam conversion matrix mentioned in step 3 is obtained by the following steps: The first step is to determine the angular range of all radar sources. Uniform division Angles in each direction ; The second step is to apply the horizontal polarization array... The directional steering vectors are combined into a beam conversion matrix for the horizontal polarization array: , for 3D matrix; The third step is to apply the horizontal polarization array... The directional steering vectors are combined into a beam conversion matrix for the vertical polarization array: , for 3D matrix.

5. The forming method according to claim 4, characterized in that, The whitening pretreatment steps described in step 4 are as follows: The first step is to convert the beam conversion matrix of the horizontal polarization array. Whitening preprocessing is performed to obtain the preprocessed beam conversion matrix. , for 3D matrix This represents the matrix conjugate transpose operation; The second step is to adjust the beam conversion matrix of the vertical polarization array. Whitening preprocessing is performed to obtain the preprocessed beam conversion matrix. , for 3D matrix.

6. The forming method according to claim 5, characterized in that, The steps in step 4 for converting the received signals and steering vectors from the horizontally polarized and vertically polarized array element domains to the beam domain are as follows: The first step is to receive signals using a horizontally polarized array. The signal is converted to the beam domain to obtain the horizontally polarized array received signal in the beam domain. , express A three-dimensional matrix, where the value of L is equal to the total number of snapshots; The second step is to guide the horizontal polarization to the vector. Transform to the beam domain to obtain the beam domain horizontal polarization steering vector. , express 3D guiding vector; The third step is to receive the signal using the vertical polarization array. The signal is converted to the beam domain to obtain the beam domain vertical polarization array received signal. , express A three-dimensional matrix, where the value of L is equal to the total number of snapshots; The fourth step is to guide the vertical polarization vector. Transform to the beam domain to obtain the beam domain vertical polarization steering vector. , express Dimensional guiding vector.

7. The forming method according to claim 6, characterized in that, The steps in step 5 regarding the reassembly of the beam domain array received signal and the beam domain steering vector are as follows: The first step is to receive the signal using a horizontally polarized array in the beam domain. and vertical polarization array receive signal Dual-polarized arrays that can be combined into a beam domain to receive signals : ； in, Represented as 3D matrix The value is less than , The value is equal to the sum of the number of elements in the horizontal polarization array and the number of elements in the vertical polarization array; The second step is to guide the horizontal polarization of the beam domain to a vector. By merging with the zero vector, the beam-domain horizontal polarization hybrid steering vector is obtained. : ； in, express The zero vector of dimension, The value of is equal to , express The guide vector; The third step is to convert the zero vector and the beam domain vertical polarization guide vector. By merging, a beam-domain vertical polarization hybrid steering vector is obtained. : ； in, express The zero vector of dimension, The value of is equal to , express The guide vector.

8. The forming method according to claim 7, characterized in that, Step 6, which describes performing different adaptive weighting processes on the dual-polarized array received signal in the beam domain, refers to first utilizing the dual-polarized array received signal in the beam domain... Estimating the joint covariance matrix of polarization domain and spatial domain : ； in, express The joint covariance matrix of the beam domain, polarization domain, and spatial domain. express Selection matrix for interference snapshots of dual-polarized arrays in the beam domain. Representation matrix The column number, whose value represents the signal received from the dual-polarized array in the beam domain. Total number of snapshots The number of data points to be extracted, and its range is: The selected snapshot number is not equal to the snapshot number of the target.

9. The forming method according to claim 8, characterized in that, The steps of the adaptive weighting process are as follows: The first step is to utilize the joint covariance matrix of the polarization domain and the spatial domain. and horizontal polarization hybrid steering vector in the beam domain Calculate the horizontal polarization adaptive weight vector in the beam domain. , for 3D vector, to obtain horizontal polarization output data The instantaneous power of the signal is obtained by taking the square of the modulus. ; The second step utilizes the joint covariance matrix of the polarization domain and the spatial domain. Vertical polarization hybrid steering vector in the beam domain Calculate the vertical polarization adaptive weight vector in the beam domain. , for 3D vector, to obtain vertical polarization output data The instantaneous power of the signal is obtained by taking the square of the modulus. .

10. The forming method according to claim 1, characterized in that, The output SINR of the horizontal and vertical polarization output data mentioned in step 7 is obtained by the following steps: The first step is to determine the instantaneous power of the horizontally polarized output signal. Calculate the SINR output and output the horizontal polarization data. The form of decibel is: ； in, This represents the power of the horizontally polarized output target signal. This represents the sum of interference power and noise power in the horizontally polarized output data; The second step is to determine the instantaneous power of the vertically polarized output signal. Calculate the output SINR and vertical polarization output data. The form of decibel is: ； in, This indicates the power of the vertically polarized output target signal. This represents the sum of interference power and noise power in the vertically polarized output data.

Images