Cognitive fda-mimo radar deception jamming suppression target detection method

Abstract

The application discloses a cognitive FDA-MIMO radar deception jamming suppression target detection method, comprising the following steps: acquiring a nonlinear motion target signal received on a range gate; using a nonlinear UKF filter to predict the state and weight of a sampling point set, and estimating the state of the nonlinear motion target; performing distance gate energy solving according to the nonlinear motion target signal received on the range gate, acquiring an energy gradient sequence, extracting a range gate serial number set, and respectively constructing a nonlinear motion target covariance matrix and an interference covariance matrix; suppressing the interference noise covariance matrix, and acquiring a nonlinear motion target signal power vector on all range gates; detecting the nonlinear motion target signal power vector on all range gates, and acquiring an updated nonlinear motion target signal; and tracking the updated nonlinear motion target signal. The application can improve the detection performance.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of radar signal processing, and particularly relates to a cognitive FDA-MIMO radar deception jamming suppression target detection method. BACKGROUND

[0002] With the rapid development of modern electronic countermeasure technology, the challenge of target detection and tracking is increasingly severe. Various intentional interference seriously affects the detection performance of traditional phased array radars, and especially the existence of main lobe deception jamming brings a severe challenge to traditional phased array radars.

[0003] In the related art, the non-cognitive FDA-MIMO radar parameter estimation cannot dynamically adjust the radar parameters, and cannot achieve optimal detection, parameter estimation and tracking performance; the training samples of the traditional interference plus noise covariance matrix do not consider eliminating target signals; in addition, the target prior information provided by the tracker to the radar system itself has errors, and the steering vector constructed using the prior information provided by the tracker itself is biased, which will cause signal cancellation phenomenon under this non-ideal condition, and further cause low signal to interference plus noise ratio (SINR) and reduce the detection probability of the target; the traditional data-independent sample selection method selects training samples from the vicinity of the to-be-detected unit, and then processes them by distance gate. However, this method cannot effectively suppress interference, and when the to-be-detected unit exists deception jamming, the interference will be retained as a target, which will affect the detection effect.

[0004] Therefore, it is urgent to improve the defects existing in the prior art. SUMMARY

[0005] In order to solve the above problems existing in the prior art, the application provides a cognitive FDA-MIMO radar deception jamming suppression target detection method. The technical problems to be solved by the application are solved through the following technical solutions:

[0006] In the first aspect, the application provides a cognitive FDA-MIMO radar deception jamming suppression target detection method, comprising:

[0007] Constructing a signal model of a non-linear motion target and a main lobe deception trajectory jamming under a FDA-MIMO radar system, and obtaining a non-linear motion target signal received on a range gate;

[0008] Performing Sigma point sampling on the non-linear motion target signal through an unscented transformation, obtaining a sampling point set, predicting the state and weight of the sampling point set using a non-linear UKF filter, and estimating the state of the non-linear motion target;

[0009] According to the nonlinear motion target signal received by the distance gate, distance gate energy solving is performed to obtain an energy gradient sequence, according to the energy gradient sequence, a distance gate serial number set is extracted, and according to the distance gate serial number set, a nonlinear motion target covariance matrix and an interference covariance matrix are respectively constructed;

[0010] An adaptive minimum mean square distortionless response beamformer is used to suppress the interference noise covariance matrix to obtain a nonlinear motion target signal power vector on all distance gates;

[0011] A cascaded constant false alarm detector is used to detect the nonlinear motion target signal power vector on all distance gates to obtain an updated nonlinear motion target signal;

[0012] The updated nonlinear motion target signal is tracked again.

[0013] Optionally, the nonlinear motion target signal received by the distance gate comprises:

[0014] The transmission signal of the i th transmission array element is obtained; A preset far-field point nonlinear motion target is obtained, and a first echo signal reflected by the i th receiving array element after being reflected by the far-field point nonlinear motion target is obtained;

[0015] After the first echo signal is mixed and matched filtered, a target signal received by the i th distance gate is obtained;

[0016] A preset false target is obtained, and a second echo signal reflected by the i th receiving array element after being reflected by the false target is obtained; After the second echo signal is mixed and matched filtered, an interference signal received by the i th distance gate is obtained;

[0017] According to the target signal and the interference signal, a nonlinear motion target signal received by the distance gate is obtained.

[0018] Optionally, the state of the sampling point set is predicted, including target state kernel prediction and target state covariance matrix prediction; wherein, The expression of the target state kernel prediction is:

[0019]

[0020]

[0021]

[0022]

[0023] is the i th​​​​​​​​ a coherent processing time period, a state of a a sample in a coherent processing time period, a state of a a weight of a a sample in a coherent processing time period, a weight of a a dimension of a state of a nonlinear moving target, a probability of state transition, a process noise subject to a covariance matrix with a mean of 0 and a nonlinear state equation function;

[0024] An expression of a target state covariance matrix prediction is as follows:

[0025]

[0026] wherein, a covariance matrix of the process noise, a transpose symbol.

[0027] Optionally, the nonlinear moving target covariance matrix and the interference covariance matrix are constructed, comprising:

[0028] estimating a nonlinear moving target signal energy received on a distance gate , obtaining a sequence thereof, and obtaining a nonlinear moving target signal energy gradient sequence after denoising and sorting;

[0029] a preset threshold, extracting a distance gate sequence number of a nonlinear moving target signal greater than the preset threshold in the nonlinear moving target signal energy gradient sequence, and obtaining a distance gate sequence number set;

[0030] obtaining a predicted distance gate sequence number of the nonlinear moving target according to a state of the nonlinear moving target, comparing the predicted distance gate sequence number of the nonlinear moving target with elements in the distance gate sequence number set, and respectively obtaining a distance gate sequence number set of interference and a distance gate sequence number set of noise;

[0031] respectively obtaining nonlinear moving target and noise sample data, interference and noise sample data, and noise sample data according to the distance gate sequence number set, the distance gate sequence number set of interference, and the distance gate sequence number set of noise;

[0032] respectively obtaining nonlinear moving target and noise covariance matrices, interference and noise covariance matrices, and noise covariance matrices according to the nonlinear moving target and noise sample data, the interference and noise sample data, and the noise sample data;

[0033] ​​Based on the nonlinear moving target and noise covariance matrix, the disturbance and noise covariance matrix, and the noise covariance matrix, the nonlinear moving target covariance matrix and the disturbance covariance matrix are obtained.

[0034] Optionally, obtain the power vectors of nonlinear moving target signals on all range gates, including:

[0035] Based on the nonlinear moving target covariance matrix and the interference covariance matrix, the signal-to-interference-plus-noise ratio is obtained;

[0036] The maximum weight vector in maximizing the signal-to-interference-plus-noise ratio is solved using an adaptive minimum mean square distortion-free response beamformer.

[0037] Based on maximizing the weight vector, the power of the nonlinear moving target signal on the range gate is obtained sequentially, and the power vector of the nonlinear moving target on all range gates is obtained.

[0038] Optionally, the construction process of an adaptive minimum mean square distortion-free response beamformer includes:

[0039] An adaptive minimum mean square distortion-free response beamformer is constructed based on the estimated nonlinear target distance, angle, and interference plus noise covariance matrix in the nonlinear moving target state.

[0040] Optionally, a cascaded constant false alarm rate (CFAR) detector is used to detect the power vector of the nonlinear moving target signal on all range gates to obtain updated nonlinear moving target signals, including:

[0041] The power vector of the nonlinear moving target signal on all distance gates is detected using a preset first-level constant false alarm detector to determine the distance gate number and radial distance measurement of the nonlinear moving target.

[0042] Preset virtual guide vector, obtain the first The signal power at each angle resolution unit is obtained, and the signal power vector at the angle resolution unit is acquired.

[0043] The angle detection information of the nonlinear moving target is obtained by using a preset second-level constant false alarm detector;

[0044] The current measurement value of a nonlinear moving target is obtained by using a cascaded constant false alarm rate detector.

[0045] Based on the current measurement value of the nonlinear moving target, obtain the updated nonlinear moving target signal.

[0046] Optionally, the process of acquiring updated nonlinear moving target signals includes:

[0047] Solve for the predicted measurements of the sample point set;

[0048] Based on the predicted measurement values ​​of the sampling point set and the measurement values ​​of the nonlinear moving target at the current moment, the measurement covariance matrix is ​​obtained;

[0049] The state covariance matrix is ​​obtained based on the predicted measurements of the sampling point set and the current measurement of the nonlinear moving target.

[0050] Update the nonlinear moving target state and nonlinear moving target state covariance matrix at the current moment based on the measurement covariance matrix and the state covariance matrix.

[0051] The beneficial effects of this invention are:

[0052] This invention provides a cognitive FDA-MIMO radar deception jamming suppression and target detection method. First, a cognitive FDA-MIMO radar deception jamming suppression and target detection system is constructed. Second, an adaptive sample selection method based on tracking prior is proposed to select non-homogeneous jamming samples, thereby obtaining an accurate jamming + noise covariance matrix. Third, based on the ASS algorithm, two adaptive beamformers are proposed—an ASS-MVDR adaptive beamformer based on maximum output SINR and a Robust-ASS adaptive beamformer—to achieve jamming suppression. Finally, the latest target measurement is obtained through a cascaded CFAR detector and provided to the tracker to achieve target state update and target state prediction for the next CPI, thus forming a closed-loop cognitive system.

[0053] The present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0054] Figure 1 This is a flowchart of a cognitive FDA-MIMO radar deception jamming suppression target detection method provided in an embodiment of the present invention;

[0055] Figure 2 This is a schematic diagram of a signal model for nonlinear moving targets and main lobe deception trajectory interference provided in an embodiment of the present invention;

[0056] Figure 3 This is a flowchart illustrating the construction of the nonlinear moving target covariance matrix and the disturbance covariance matrix provided in this embodiment of the invention.

[0057] Figure 4 This is a schematic diagram of UKF target location prediction sampling and 3-sigma confidence ellipse provided in an embodiment of the present invention;

[0058] Figure 5 This is a schematic diagram of UKF target velocity prediction sampling and 3-sigma confidence ellipse provided in an embodiment of the present invention;

[0059] Figure 6(a) is a schematic diagram of the spatial distribution of the target and deceptive interference at different CPI times provided in the embodiments of the present invention;

[0060] Figure 6(b) is a schematic diagram of a nonlinear motion trajectory of a target in the XY plane provided by an embodiment of the present invention;

[0061] Figure 7 This is a statistical chart of the detection probability of targets under different algorithms and different SNRs provided in the embodiments of the present invention;

[0062] Figure 8 This is a schematic diagram of the distance and angle tracking results of different algorithms provided in this embodiment of the invention at SNR=-6.8dB;

[0063] Figure 9 This is a schematic diagram of the distance-angle RMSE estimation results of different algorithms provided in the embodiments of the present invention when SNR=-6.8dB;

[0064] Figure 10 This is a comparison chart of the output SINR of different algorithms provided in the embodiments of the present invention;

[0065] Figure 11 This is a schematic diagram showing the target distance tracking results and error statistics of the MVDR-ASS algorithm under different signal-to-noise ratios provided in the embodiments of the present invention;

[0066] Figure 12 This is a schematic diagram showing the target angle tracking results and error statistics of the MVDR-ASS algorithm under different signal-to-noise ratios provided in the embodiments of the present invention;

[0067] Figure 13 This is a schematic diagram of the SINR output curve of the MVDR-ASS algorithm under different SNRs as a function of tracking steps, provided in an embodiment of the present invention.

[0068] Figure 14 This is a schematic diagram of Robust-ASS, which provides cognitive distance and angle tracking results under different SNRs according to an embodiment of the present invention.

[0069] Figure 15 This is a schematic diagram of Robust-ASS, which provides distance and angle RMSE estimation results under different SNRs according to an embodiment of the present invention.

[0070] Figure 16 This is a schematic diagram of Robust-ASS, which provides the output SINR estimation results under different SNR at different CPI times according to an embodiment of the present invention. Detailed Implementation

[0071] The present invention will be further described in detail below with reference to specific embodiments, but the implementation of the present invention is not limited thereto.

[0072] In existing technologies, Paul Antonik et al. first introduced a frequency offset much smaller than the carrier frequency between the transmitting elements of a phased array, thus proposing the concept of frequency diversity arrays. However, Paul Antonik et al. only conducted a preliminary exploration of the emission pattern characteristics of frequency diversity arrays (FDA) without in-depth research. Thanks to the unique radiation characteristics of FDA itself, it has attracted a large number of scholars to analyze and interpret its emission pattern characteristics from multiple perspectives, revealing the range-angle coupling characteristics of the FDA emission pattern. Some scholars have also designed different frequency offsets to control its emission pattern in order to obtain better target detection and tracking performance. However, this unsteady characteristic of the range-angle coupling of the FDA emission pattern has brought difficulties to parameter estimation. Therefore, it is necessary to combine multiple-input multiple-output (MIMO) radar technology to separate the transmitted waveforms of each array element at the receiver through mixing and matched filtering. Then, range-angle joint estimation, clutter suppression, and anti-mainlobe deception interference can be achieved by constructing a range-angle joint matched filter. However, the relevant literature on anti-mainlobe interference mentioned above focuses on stationary targets or single moments and does not consider target tracking and cognition methods.

[0073] In fact, cognitive radar can perceive external environmental information and feed it back to the transmitter through a closed-loop cognitive approach, forming a fully adaptive radar processing architecture with a dynamic closed loop between the receiver, transmitter, and environment. Cognitive radar can adjust the radar's transmitted waveform in real time using target and environmental information fed back from the tracker to achieve optimal target detection and tracking performance. In the field of FDA-MIMO radar cognitive technology, the earliest research focused on leveraging the low intercept advantage of MIMO radar, optimizing the transmission weight vector based on the principles of minimizing energy at the target location and maximizing the output signal-to-interference-plus-noise ratio, thus ensuring low intercept characteristics without affecting target detection performance. Basit, A., Qureshi, IM, Khan, W., et al. used a cognitive approach to achieve adaptive adjustment of frequency offset in FDA-MIMO radar, thereby focusing energy on the target area and achieving better target detection performance. R. Gui, W.-Q. Wang, Y. Pan and J. Xu, et al. introduced subarrays into the FDA-MIMO radar system. By setting the subarrays to phased array mode and the subarrays to FDA mode, the entire cognitive system simultaneously possesses the high gain characteristics of phased arrays and the additional range dimension of freedom of FDA-MIMO radar. A. Basic, W.-Q. Wang, SYNusenu, and Z. Zheng et al. adjusted the beam to achieve better target detection and tracking performance by introducing time-dependent frequency increments; it is worth noting that none of the above literatures considered interference. Recently, based on the maximum output SINR principle and MVDR beamformer, cognitive FDA-MIMO anti-interference was achieved through power allocation; B. Yang, S. Zhu, X. He, L. Lan, and X. Li et al. used a networking approach to achieve target and interference identification and proposed a typical cognitive target tracking technique based on the maximum Capon power spectrum principle; however, some studies have not considered how to select samples to construct the interference plus noise covariance matrix.

[0074] In view of this, the present invention provides a target detection method for suppressing deception jamming in cognitive FDA-MIMO radar. First, addressing the problem that traditional anti-jamming algorithms struggle to obtain prior target information and accurate jamming plus noise covariance matrix, the present invention proposes an adaptive sample selection (ASS) algorithm using target prediction information provided by the tracker to train an accurate jamming plus noise covariance matrix. Second, considering the error in target prediction information provided by the tracker, based on the ASS algorithm, the present invention proposes a minimum variance distortionless response-based adaptive sample selection (MVDR-ASS) and a robust beamformer-based adaptive sample selection (Robust-ASS) based on the maximum signal-to-interference plus noise ratio criterion (MSINR), which can effectively achieve jamming suppression.

[0075] Please see Figure 1 As shown, Figure 1 This is a flowchart of a cognitive FDA-MIMO radar deception jamming suppression target detection method provided by an embodiment of the present invention. The cognitive FDA-MIMO radar deception jamming suppression target detection method provided by the present invention includes:

[0076] S101. Construct a signal model for nonlinear moving targets and main lobe deception trajectory interference under the FDA-MIMO radar system, and obtain the nonlinear moving target signal received at the range gate.

[0077] Specifically, please see Figure 2 As shown, Figure 2 This is a schematic diagram of a signal model for nonlinear moving targets and main lobe deception trajectory interference provided in an embodiment of the present invention. In this embodiment, a signal model for nonlinear moving targets and main lobe deception trajectory interference under an FDA-MIMO radar system is first constructed, and a co-located FDA-MIMO array is set up, with the transmitter and receiver respectively consisting of... Each launch element and A uniform linear array consisting of receiving elements, with the spacing between adjacent transmitting elements being... , For wavelength, For reference frequency, Unlike traditional MIMO arrays, this embodiment uses the speed of light. At the transmitter of the FDA-MIMO array, the frequencies of the transmitted signals from adjacent transmitter elements differ by a factor of two. If the frequency of the signal transmitted by the first transmitting element is Then the adjacent launch array element The frequency of the transmitted signal is .

[0078] According to the The frequency of the transmitted signal of each transmitting element , obtain the The transmission signal of each transmitting element Its expression is:

[0079] ;

[0080] in, The duration of the pulse. To the total energy emitted, The number of elements in the transmitting array. For the first The envelope of the signal transmitted by each transmitting element. It is an exponential function. The imaginary unit, Pi For the first The frequency of the signal transmitted by each transmitting element For the transmission time, the envelopes of the transmitted signals of each transmitting element satisfy the orthogonality condition, which is expressed as:

[0081] ;

[0082] Set the total launch energy Equal to the number of elements in the launch array Then the first The transmitted signals of each transmitting element are simplified as follows:

[0083] .

[0084] Preset far-field point target After being reflected by the far-field point target, the image is obtained. The time delay of the echo signal received by each receiving element The first term on the right side of the equation is the common two-way delay. The second term is the path difference caused by the distance between the transmitting and receiving array elements; where, The initial distance of the nonlinearly moving target is . Let be the initial angle of the nonlinearly moving target. For public two-way delay, The spacing between adjacent transmitting elements. At the speed of light, It is a sine function.

[0085] Under the narrowband assumption, the first The echo signal received by each receiving element The expression is:

[0086] ;

[0087] in, The frequency of the signal transmitted by the first transmitting element. The frequency difference between the signals transmitted by adjacent transmitting elements; The frequency of the signal transmitted by the first transmitting element;

[0088] The first The echo signal received by each receiving element go through (Regarding the first) After mixing by (each transmitting array element), the mixed signal is obtained, and its expression is:

[0089] ;

[0090] The mixed signal is then subjected to matched filtering to obtain the processed signal, the expression of which is:

[0091] ;

[0092] The signal processed above is then subjected to waveform orthogonality processing to obtain:

[0093] ;

[0094] in, The target echo coefficient (complex number) includes the array element transmit gain, array element receive gain, transmit and receive, target backscattering gain, and pulse compression processing gain. It should be noted that the last two terms in the above formula can be ignored when the following condition is met:

[0095] ;

[0096] When the above conditions are met, the first The echo signal received by each receiving element It can be simplified to:

[0097] .

[0098] Assuming the target is located at the th pulse compression point The distance gate, after mixing and matched filtering, the... The signal on the distance gate can be represented as:

[0099] ;

[0100] in, For the launch guidance vector, The expressions for receiving the guide vector are as follows:

[0101] ;

[0102] ;

[0103] ;

[0104] ;

[0105] Set to transmit within each coherent processing period (CPI). A pulse, for a velocity of Nonlinear motion target at far field point Considering the Doppler frequency shift caused by the nonlinear motion of the target. After mixing and matched filtering, the first The signal received by the distance gate is:

[0106] ;

[0107] in, The Doppler steering vector, Accumulate time for coherence.

[0108] In this embodiment, considering A deceptive interference ,in, For the first The distance between the dummy target and the receiving array element. For the first The angle between the dummy target and the receiving array element For Doppler frequency shift, the first The signal from the first dummy target, after being mixed and matched filtered, is located at the... Then the distance gate, then the first The expression for the interference signal received at each distance gate is:

[0109] ;

[0110] For radar signals processed by mixing and matched filtering, the target and interference may be located at any range gate; therefore, the range gate... The received signal, i.e., the nonlinear moving target signal, can be expressed as:

[0111] ;

[0112] ;

[0113] in, For the Dirac impulse function, This is the noise term.

[0114] S102. The nonlinear moving target signal is sampled using Sigma points through unscented transformation to obtain a set of sampling points. The state and weights of the sampling point set are predicted using a nonlinear UKF filter to estimate the state of the nonlinear moving target.

[0115] Specifically, in this embodiment, a nonlinear UKF filter is used to predict the state and weights of the sample point set to obtain the target state and covariance matrix estimate.

[0116] Unscented transformations (UT transformations) are an efficient method for calculating random variables undergoing nonlinear transformations; consider nonlinear state transition functions. Using the target state during the previous coherent processing time period (CPI) Covariance Matrix ,in, and It contains the position and velocity information of the nonlinear moving target, respectively.

[0117] Sigma point set through UT transformation Sampling, its expression is:

[0118] ;

[0119] in, For the first One CPI, and The first The first CPI The state and weight of each sample, Let the dimension of the target state be . For scaling factor, parameter and This determines the distribution of Sigma points around the mean.

[0120] In this embodiment, the prediction of the state of the sampling point set includes target state kernel prediction and target state covariance matrix prediction, wherein...

[0121] The expression for the target state kernel prediction is:

[0122] ;

[0123] For the first A coherent processing time period, For the first The sampling points in the first coherent processing time period are concentrated in the first... The state of each sample For the first The sampling points in the first coherent processing time period are concentrated in the first... The weights of each sample, Let be the dimension of the nonlinear moving target state. The probability of a state transition. The covariance matrix follows a mean of 0. Process noise, It is a nonlinear state equation function;

[0124] The expression for predicting the target state covariance matrix is:

[0125] ;

[0126] in, Let be the covariance matrix of the process noise. This is the transpose symbol.

[0127] S103. Based on the nonlinear moving target signal received at the range gate, perform range gate-by-range energy calculation to obtain the energy gradient sequence. Based on the energy gradient sequence, extract the range gate number set. Based on the range gate number set, construct the nonlinear moving target covariance matrix and the interference covariance matrix respectively.

[0128] Specifically, please see Figure 3 As shown, Figure 3 This is a flowchart of constructing the nonlinear moving target covariance matrix and the interference covariance matrix provided in this embodiment of the invention. In this embodiment, an adaptive sample selection algorithm based on target prior (the target prior is obtained from the tracker's prediction value) and signal energy gradient is proposed to accurately construct the covariance matrix of interference, noise and target signal.

[0129] First, the energy gradient sequence is obtained by calculating the energy of all signals at the range gates and arranging them in descending order. Second, since the energy of the range gates with interference and targets is much greater than that of the range gates with only noise, the range gate numbers of the interference and targets can be determined by the threshold method. Then, the range gate number of the target is determined by using the target's prior information, thereby obtaining the range gate numbers of the interference and noise. Finally, the interference and noise samples are selected to accurately train the interference and noise covariance matrix as the basis for anti-interference.

[0130] After mixing and matched filtering, the first... A distance from the door The pulse data is Then the first The signal energy received at each distance gate can be estimated as follows:

[0131] ;

[0132] The signal energy sequence at all distance gates is as follows: ,in, , After sorting in descending order, the energy gradient sequence is obtained, and its expression is:

[0133] ;

[0134] in, The reordered distance gate indices have a mapping relationship that satisfies:

[0135] ;

[0136] Considering that under high signal-to-noise ratio and high interference-to-noise ratio conditions, the energy of the target and deceptive interference is much greater than the energy of the background noise, and the number of targets and interference is limited, therefore, it is possible to refer to... The last one The threshold is set by the energy of the element. The expression for obtaining the preset threshold is:

[0137] ;

[0138] in, A preset threshold factor is used to control the threshold. size.

[0139] Based on the characteristic that the energy of the target and the interference is greater than the energy of the noise, the range gate numbers that are greater than the threshold in the energy sequence can be extracted as the range gate numbers of the interference and the target.

[0140] ;

[0141] In the above formula, obtaining makes The set of distance numbers that are established It is understandable that the range gate numbers of the target and the interference signal form a set. , for The inverse mapping; it should be noted that the set The signal at the mid-range gate has added noise; the gate number containing only noise can be represented as follows: In the cognitive structure, since the tracker can continuously provide the system with target prediction information, this prediction information can be used to predict the distance gate number of the target; assuming the first... The forecast status of the CPI target is as follows: The predicted radial distance and angle of the target are:

[0142] ;

[0143] The predicted target distance gate number is:

[0144] ;

[0145] Where ceil is the floor function. The distance from the door width, The signal bandwidth is used to compare the target predicted range gate number. and set Middle element, distance Minimum distance gate number This is the distance gate of the real target; therefore, the distance gate number of the interference can be represented as... For ease of remembering the distance gate numbers where interference occurs, the set of gate numbers is as follows: The set of distance gate indices containing only noise is .

[0146] Acquire target and noise sample data .

[0147] Interference and noise sample data:

[0148] ;

[0149] Noise sample data:

[0150] ;

[0151] in, The data is arranged into a matrix according to the distance gate number.

[0152] Therefore, the target and noise covariance matrices can be obtained as follows:

[0153] ;

[0154] Interference and noise covariance matrix:

[0155] ;

[0156] Noise covariance matrix:

[0157] ;

[0158] To avoid problems caused by too many distance doors To address the issue of excessively high dimensionality, the noise covariance matrix can also be solved using the following formula:

[0159] ;

[0160] The precise target signal covariance matrix is ​​then: The interference covariance matrix is .

[0161] S104. Use an adaptive minimum mean square distortion-free response beamformer to suppress the interference plus noise covariance matrix and obtain the power vector of nonlinear moving target signals on all range gates.

[0162] Specifically, this embodiment proposes an adaptive sample selection (MVDR-ASS) beamformer based on the maximum output signal-to-interference plus noise ratio criterion (MSINR).

[0163] Based on the adaptive sample selection method described above, the target signal covariance matrix and the interference-plus-noise covariance matrix can be accurately constructed. The signal-to-interference-plus-noise ratio (SINR) can then be defined as:

[0164] ;

[0165] in, The adaptive weight vector can be solved using an MVDR beamformer:

[0166] ;

[0167] in, Distance and angle Dependent virtual guide vector, distance and angle The target distance and angle are predicted by the tracker. It should be noted that the tracker's predicted target distance and angle contain errors. Directly using the calculated weight vector would be biased. Therefore, the optimal weight vector is solved by maximizing the output SINR.

[0168] ;

[0169] in, Based on the weight vector formula and the target's preset distance and angle information Please provide a solution. and These represent the ranges of values ​​for the target's preset distance and angle, respectively, with the center of the range being the tracker's predicted value. .

[0170] In this embodiment, the distance range radius is 3 times the distance gate width, and the angle range radius is 3 times the angle search unit.

[0171] In one embodiment of the present invention, the Robust-ASS beamformer based on tracking prior is an MVDR-ASS beamformer based on MSINR. Using the tracker, the predicted target state value for the next CPI can be obtained. Therefore, based on the predicted target distance, angle, and interference plus noise covariance matrix, a Robust-ASS beamformer can be constructed, the expression of which is:

[0172] ;

[0173] Unlike existing technologies, the interference plus noise covariance matrix in the adaptive beamformer provided in this embodiment has already removed the target signal, rather than being a sampled covariance matrix containing the target signal component. Based on the constructed robust adaptive beamformer, the optimal weight vector can be solved to achieve interference suppression. To find the imaginary part, Define the boundary (a known constant) of the error of the virtual steering vector, satisfying:

[0174] ;

[0175] in, and The first The distance and angle from the true CPI target. A virtual guide vector for the real target.

[0176] S105. Use cascaded constant false alarm rate detectors (CFAR) to detect the power vector of the nonlinear moving target signal on all distance gates and obtain the updated nonlinear moving target signal.

[0177] Specifically, in this embodiment, cascaded CFAR is used for target measurement extraction.

[0178] No. The optimal weight vector obtained from CPI training is The signal power at each distance gate can be solved sequentially. For the th distance gate... After using a distance gate and matched filtering with weight vectors, the signal power is:

[0179] ;

[0180] in, Represented as ;

[0181] The signal power vectors at all distance gates are then:

[0182] ;

[0183] Therefore, range direction detection can be performed by designing a first-stage unit averaged CFAR detector. The threshold of the detector on the unit to be detected is:

[0184] ;

[0185] in, The preset false alarm probability, Number of reference units The average power of the reference cell.

[0186] The range gate of the target can be determined using a Level 1 CFAR detector. Measurement of radial distance .

[0187] At this point, a virtual guide vector can be designed. Solve the first... Signal power on each angle resolution unit:

[0188] ;

[0189] The signal power vectors at all angular units are then:

[0190] ;

[0191] in, Given the number of angle units, a second-stage CFAR detector can be designed to obtain the target angle detection information as follows:

[0192] ;

[0193] in, For the size of the angle unit, For CFAR detection, the angular element number is obtained by the detector. Therefore, the current target measurement value can be obtained by cascading CFAR detectors.

[0194] ;

[0195] The detection values ​​at the current moment can be provided to the tracking system to update the target state at that moment. Furthermore, the above analysis shows that the detection errors for distance and angle are respectively:

[0196] and ;

[0197] Therefore, the tracking error covariance matrix is:

[0198] ;

[0199] in, This is the vector diagonalization operator.

[0200] Based on the current measurement value of the nonlinear moving target, obtain the updated nonlinear moving target signal.

[0201] Solve for the predicted measurements of the sample point set;

[0202] No. The predicted state of each sample is The corresponding predicted measurement value is:

[0203] ;

[0204] The predicted measurement of the target state at the current moment is:

[0205] ;

[0206] Based on the predicted measurements from the sampling point set and the current measurement values ​​of the nonlinearly moving target, the measurement covariance matrix is ​​obtained, and its expression is:

[0207] ;

[0208] Based on the predicted measurements from the sampling point set and the current measurement of the nonlinear moving target, the state covariance matrix is ​​obtained, and its expression is:

[0209] ;

[0210] Update the nonlinear moving target state and nonlinear moving target state covariance matrix at the current moment based on the measurement covariance matrix and the state covariance matrix;

[0211] The Kalman gain is then... Then the first The CPI target state and target state covariance matrix are updated as follows:

[0212] ;

[0213] ;

[0214] Please see Figure 4 and Figure 5 As shown, Figure 4 This is a schematic diagram of UKF target location prediction sampling and 3-sigma confidence ellipse provided in an embodiment of the present invention; Figure 5 This is a schematic diagram of UKF target velocity prediction sampling and 3-sigma confidence ellipse provided in an embodiment of the present invention; when SNR=-6dB, the prediction update of the distribution of the 4th CPI Sigma sampling point, the prior 3-Sigma confidence ellipse, the predicted 3-Sigma confidence ellipse, and the updated 3-Sigma confidence ellipse; the prior target position and velocity refer to the target position and velocity information estimated by the previous CPI tracker. It can be seen that the sampling points are concentrated in the 3-Sigma confidence ellipse, and the sampling points are mostly concentrated near the true value; in addition, after prediction and update, the 3-Sigma confidence ellipse of the target position and velocity gradually converges to the true target position.

[0215] S; 106. Then, track the updated nonlinear motion target signal.

[0216] Specifically, in this embodiment, the updated target state and covariance matrix can be used to predict the target state and covariance matrix of the next CPI, and then the predicted value is provided to the signal processing system to continue the adaptive sample selection and anti-interference of the next CPI, thereby forming a closed-loop cognitive system.

[0217] In summary, this invention provides a cognitive FDA-MIMO radar deception jamming suppression and target detection method. First, a cognitive FDA-MIMO radar deception jamming suppression and target detection system is constructed. Second, an adaptive sample selection method based on tracking prior is proposed to select non-homogeneous jamming samples, thereby obtaining an accurate jamming + noise covariance matrix. Third, based on the ASS algorithm, two adaptive beamformers are proposed—an ASS-MVDR adaptive beamformer based on maximum output SINR and a Robust-ASS adaptive beamformer—to achieve jamming suppression. Finally, the latest target measurement is obtained through a cascaded CFAR detector and provided to the tracker to achieve target state updates and target state prediction for the next CPI, thus forming a closed-loop cognitive system.

[0218] In an optional embodiment of the present invention, the effectiveness of the cognitive FDA-MIMO radar deception jamming suppression and target detection method is verified through simulation experiments.

[0219] I. Simulation Parameter Settings

[0220] The radar is located at the origin of the coordinate system. The speed of light is The carrier frequency of the signal transmitted by each array element is , wavelength is The frequency shift between adjacent array elements is Array element spacing The pulse repetition frequency is signal bandwidth Distance from door size Maximum unambiguous distance The number of transmitting and receiving array elements is The coherent processing time is 500 represents the number of accumulated pulses within one CPI. During CFAR detection, the search window sizes for the range and azimuth directions are respectively... and Therefore, the measurement noise covariance matrix is:

[0221] ;

[0222] In the simulation experiment, the target tracking time was 60 CPI, and the tracking interval was... The initial value of the target position is... The initial velocity of the target is angular velocity of rotation The initial target state covariance matrix is .

[0223] The target motion follows a nonlinear motion model:

[0224] ;

[0225] in, To follow the pattern of mean 0 and covariance , Gaussian white noise, This represents the standard deviation of the process noise.

[0226] ;

[0227] Here is the nonlinear state transition matrix:

[0228] ;

[0229] During the detection process, the false alarm probability of the CFAR detector was set to 10⁻⁶, the distance and angle protection units were set to 10, and the number of reference units was set to 20.

[0230] Please refer to Figures 6(a) and 6(b). Figure 6(a) is a schematic diagram of the spatial distribution of the target and deceptive interference at different CPI times provided in this embodiment of the invention. Figure 6(b) is a schematic diagram of the nonlinear motion trajectory of the target in the XY plane provided in this embodiment of the invention. There are two main lobe deceptive interferences, located at the 200th and 900th range gates respectively, with the angle of the main lobe deceptive interference being the same as the angle of the real target. There are three side lobe deceptive interferences, located at the 150th, 250th, and 800th range gates respectively, with the angle of the side lobe deceptive interference being different from the target angle. The upper part follows a pseudo-random distribution with a noise-to-interference ratio of 20dB.

[0231] II. Simulation Content

[0232] Experiments were conducted to verify the target detection and tracking performance of cognitive FDA-MIMO radar using different algorithms.

[0233] First, for nonlinearly moving targets, Sigma point sampling is performed using Unscented Transform (UT). A nonlinear UKF filter is then used to predict the state and weights of the sampled point set to estimate the state of the nonlinearly moving target, and the predicted target parameters are provided to the radar signal processing system. Second, a cognitive sample selection method is proposed to accurately construct the covariance matrix of interference, noise, and target signals. Based on this, a cognitive robust adaptive beamformer and an adaptive MVDR beamformer based on the maximum output signal-to-interference-plus-noise ratio (MINR) are proposed for interference suppression. A CFAR detector is used to detect the target parameters after interference suppression. Finally, the latest target information and the UKF filter are used to achieve target tracking.

[0234] (1) Experiment 1

[0235] Please see Figure 7 As shown, Figure 7 This is a statistical chart showing the detection probability of targets under different algorithms and different SNRs provided in the embodiments of the present invention. Figure 7In this invention, MVDR-ASS is a cognitive interference suppression and target detection algorithm based on the MVDR-ASS beamformer; Robust-ASS represents a cognitive interference suppression and target detection algorithm based on the Robust-ASS beamformer; DDD refers to the direct data domain method for training the interference and noise covariance matrix. The sample covariance matrix obtained using DDD contains the target signal, therefore the detection performance of Robust-DDD and MVDR-DDD algorithms is inferior to that of Robust-ASS and MVDR-ASS algorithms. Under the same SNR, the detection probability of the Robust algorithm is greater than that of MVDR because the maximum output SINR principle, when determining the optimal weights, inevitably introduces errors due to the search intervals for distance and angle. Figure 7 The results show that the detection probability of the target can be improved by using the cognitive strategy proposed in this invention for adaptive sample selection.

[0236] Please see Figure 8 As shown, Figure 8 This is a schematic diagram illustrating the distance and angle tracking results of different algorithms provided in this embodiment of the invention at an SNR of -6.8dB. Figure 8 It can be seen that, after adopting the adaptive sample selection strategy proposed in this invention, the target tracking performance of Robust-ASS and MVDR-ASS cognitive anti-interference and target detection algorithms is far superior to the other two algorithms. This is because the interference plus noise covariance matrix of the other two algorithms is replaced by the sampling covariance matrix of the direct data domain. The sampling covariance matrix of the direct data domain contains target information, leading to the failure of interference suppression, resulting in target detection failure and tracking divergence. Furthermore, the proposed Robust-ASS algorithm has better tracking performance than the proposed MVDR-ASS algorithm because the detection probability of the Robust-ASS algorithm is 1 at SNR=-6.8dB, while the detection probability of the MVDR-ASS algorithm is less than 1. Random missed detections cause the tracking performance of the MVDR-ASS algorithm to be inferior to that of the Robust-ASS algorithm. Please refer to [link to relevant documentation]. Figure 9 As shown, Figure 9 This is a schematic diagram illustrating the distance-angle RMSE estimation results of different algorithms provided in this embodiment of the invention at SNR=-6.8dB, combined with... Figure 8 and Figure 9 As shown, the tracking performance of Robust adaptive beamformer is better than that of MVDR beamformer. This is because Robust adaptive beamformer has a better tolerance for errors by introducing a weight vector uncertainty set.

[0237] Please see Figure 10 As shown, Figure 10This is a comparison chart of the output SINR of different algorithms provided in the embodiments of the present invention. Figure 10 It can be seen that CFDA-MIMO DISTDS based on MVDR-ASS and Robust-ASS algorithms can achieve relatively ideal interference suppression performance at different CPIs. However, the output SINR of the MVDR-ASS algorithm is slightly lower than that of Robust-ASS. (Comparison) Figure 7 This is because the Robust-ASS algorithm has a detection probability of 1 at SNR = -6.8dB, while the MVDR-ASS algorithm has a detection probability of 0.99. Furthermore, the output SINR of CFDA-MIMO DISTDS based on MVDR-DDD and Robust-DDD algorithms gradually decreases with each tracking step iteration. This is because the detection probabilities of MVDR-DDD and Robust-DDD are 0.18 and 0.8, respectively. If the target is not detected, it will lead to a large error in the prior information of the target input to the signal processing system, resulting in a large error in the optimal weight vector obtained. Therefore, with the iteration of the algorithm, the interference suppression performance gradually degrades, and the output SINR gradually decreases.

[0238] (2) Experiment 2

[0239] Please see Figure 11 and Figure 12 As shown, Figure 11 This is a schematic diagram illustrating the target distance tracking results and error statistics of the proposed MVDR-ASS algorithm under different signal-to-noise ratios provided in this embodiment of the invention. Figure 12 This is a schematic diagram illustrating the target angle tracking results and error statistics of the proposed MVDR-ASS algorithm under different signal-to-noise ratios provided in this embodiment of the invention. Figure 7 It can be seen that when SNR=-6, -6.9 and -7dB, the target detection probabilities in the MVDR-ASS algorithm are 1, 0.94 and 0.51, respectively. Figure 11 and Figure 12 It can be seen that as the SNR decreases, the probability of target detection decreases, and the tracking performance also gradually declines.

[0240] Please see Figure 13 As shown, Figure 13This is a schematic diagram illustrating the change curve of the SINR output of the MVDR-ASS algorithm with tracking steps under different SNR values ​​provided in this embodiment of the invention. When SNR = -6dB, the detection probability is 1, and the target can be accurately detected at any CPI. Therefore, stable tracking can be achieved, and relatively accurate prediction information can be provided to the radar system to suppress interference. Thus, the SINR output at different CPI values ​​is the highest and remains stable. When SNR = -6.9dB, the detection probability is 0.94. If the target is not detected at any CPI, it will lead to a large error in the tracking result. At this time, the large deviation of the tracker's prediction value will cause the interference suppression performance to deteriorate sharply, resulting in detection errors. Therefore, once the phenomenon of not detecting the target occurs, the tracking performance and interference suppression performance of the MVDR-ASS algorithm will degrade sharply. When SNR = -7dB, the target detection probability of the MVDR-ASS algorithm is only 0.51, so its output SINR degrades sharply with the iteration of tracking steps. Comparing the output SINR change curves for SNR=-7dB and SNR=-6.9dB, it can be seen that the degradation of output SINR becomes more severe as SNR decreases.

[0241] (3) Experiment 3

[0242] Experiment 3 primarily verifies the performance of the cognitive interference suppression and target detection algorithms based on the Robust-ASS beamformer proposed in this invention for interference suppression and target tracking. Please refer to [link / reference]. Figure 14~Figure 16 As shown, Figure 14 This is a schematic diagram of Robust-ASS, which provides cognitive distance and angle tracking results under different SNRs according to an embodiment of the present invention. Figure 15 This is a schematic diagram of Robust-ASS, representing distance and angle RMSE estimation results under different SNRs provided in this embodiment of the invention. Figure 16 This is a schematic diagram of Robust-ASS, representing the output SINR estimation results under different SNR at different CPI times according to an embodiment of the present invention. (Observation) Figure 14-16 The following conclusions can be drawn:

[0243] ① Based on stable detection (detection probability of 1, SNR=-6dB), the Robust-ASS algorithm can maintain stable tracking;

[0244] ② As the SNR decreases, especially as the detection probability decreases, the tracking results of the Robust-ASS algorithm will diverge with each tracking step iteration. This is because once a target is not detected, the prior information provided by the tracker to the system will have a large error, resulting in a large error in the optimal weights obtained by the beamformer, a decrease in interference suppression performance, and a decrease in target detection performance. This vicious cycle leads to tracking divergence.

[0245] ③ As the SNR decreases, the output SINR will gradually decrease with each tracking step iteration. The smaller the SNR, the earlier the interference suppression performance degrades, and the faster the output SINR decreases.

[0246] It should be noted that, in this document, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations are intended to cover non-exclusive inclusion, such that an article or device comprising a list of elements includes not only those elements but also other elements not expressly listed. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the article or device comprising said element. Terms such as "connected" or "linked" are not limited to physical or mechanical connections but can include electrical connections, whether direct or indirect. The orientations or positional relationships indicated by terms such as "upper," "lower," "left," and "right" are based on the orientations or positional relationships shown in the accompanying drawings and are used only for the convenience of describing the invention and for simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be construed as limiting the invention.

[0247] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the present invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features or characteristics described may be combined in any suitable manner in one or more embodiments or examples. In addition, those skilled in the art can combine and integrate the different embodiments or examples described in this specification.

[0248] The above description, in conjunction with specific preferred embodiments, provides a further detailed explanation of the present invention. It should not be construed that the specific implementation of the present invention is limited to these descriptions. For those skilled in the art, various simple deductions or substitutions can be made without departing from the concept of the present invention, and all such modifications and substitutions should be considered within the scope of protection of the present invention.

Claims

1. A method for cognitive FDA-MIMO radar deception jamming suppression target detection, characterized in that, The application relates to a method for tracking a nonlinear motion target in an FDA-MIMO radar system. The method comprises the following steps: a signal model of a nonlinear motion target and a main lobe deceptive trajectory jamming in the FDA-MIMO radar system is established, and a nonlinear motion target signal received in a distance gate is obtained; Sigma point sampling is performed on the nonlinear motion target signal through an unscented transformation, a sampling point set is obtained, a state and a weight of the sampling point set are predicted by using a nonlinear UKF filter, and the state of the nonlinear motion target is estimated; energy in each distance gate is solved according to the nonlinear motion target signal received in the distance gate, an energy gradient sequence is obtained, a distance gate serial number set is extracted according to the energy gradient sequence, and a nonlinear motion target covariance matrix and a jamming covariance matrix are respectively established according to the distance gate serial number set; an adaptive least mean square distortionless response beamformer is used to suppress the jamming and noise covariance matrix, and a nonlinear motion target signal power vector in all distance gates is obtained; a cascaded constant false alarm detector is used to detect the nonlinear motion target signal power vector in all distance gates, and an updated nonlinear motion target signal is obtained; 2. The cognitive FDA-MIMO radar deception jamming rejection target detection method of claim 1, wherein, and the updated nonlinear motion target signal is tracked. acquiring a transmission signal of a first transmission element; The preset far-field point nonlinear motion target obtains the first echo signal received by the first receiving element after being reflected by the far-field point nonlinear motion target. The preset far-field point nonlinear motion target obtains the first echo signal received by the first receiving element after being reflected by the far-field point nonlinear motion target. The first echo signal is mixed and matched filtered to obtain a target signal received on a first distance gate ​ The preset false target is used to obtain a second echo signal received by the first receiving element after being reflected by the false target. The preset false target is used to obtain a second echo signal received by the first receiving element after being reflected by the false target. The second echo signal is mixed and matched filtered to obtain the first interference signal received on the first distance gate. The distance gate is obtained according to the target signal and the interference signal The received non-linear motion target signal is connected.

3. The cognitive FDA-MIMO radar deception jamming rejection target detection method of claim 1, wherein, The method for obtaining the nonlinear motion target signal received in the distance gate comprises the following steps: the state of the sampling point set is predicted, including target state kernel prediction and target state covariance matrix prediction; wherein, ； wherein, is the first coherent processing interval, is the first coherent processing interval, is the first coherent processing interval, is the state of the first sample in the set of samples in the first coherent processing interval, is the state of the first sample in the set of samples in the first coherent processing interval, is the weight of the first sample in the set of samples in the first coherent processing interval, is the weight of the first sample in the set of samples in the first coherent processing interval, is the dimension of the state of the nonlinear moving target, is the probability of state transition, is the process noise subject to a mean of 0 and a covariance matrix of is the process noise subject to a mean of 0 and a covariance matrix of is the process noise subject to a mean of 0 and a covariance matrix of is the nonlinear state equation function; an expression of the target state kernel prediction is as follows: ； wherein is a covariance matrix of the process noise, is a transpose symbol.

4. The cognitive FDA-MIMO radar deception jamming rejection target detection method of claim 1, wherein, an expression of the target state covariance matrix prediction is as follows: Distance gate The received nonlinear motion target signal energy is estimated, and its sequence is obtained. After denoising and sorting, the nonlinear motion target signal energy gradient sequence is obtained. the method for establishing the nonlinear motion target covariance matrix and the jamming covariance matrix comprises the following steps: a preset threshold is set, a serial number of a distance gate where a nonlinear motion target signal greater than the preset threshold in the nonlinear motion target signal energy gradient sequence is extracted, and a distance gate serial number set is obtained; according to the state of the nonlinear motion target, a predicted distance gate serial number of the nonlinear motion target is obtained, the predicted distance gate serial number of the nonlinear motion target is compared with elements in the distance gate serial number set, and a distance gate serial number set where jamming is located and a distance gate serial number set where noise is located are respectively obtained; according to the distance gate serial number set, the distance gate serial number set where jamming is located and the distance gate serial number set where noise is located, nonlinear motion target and noise sample data, jamming and noise sample data and noise sample data are respectively obtained; according to the nonlinear motion target and noise sample data, the jamming and noise sample data and the noise sample data, nonlinear motion target and noise covariance matrix, jamming and noise covariance matrix and noise covariance matrix are obtained; 5. The cognitive FDA-MIMO radar deception jamming rejection target detection method of claim 1, wherein, according to the nonlinear motion target and noise covariance matrix, the jamming and noise covariance matrix and the noise covariance matrix, nonlinear motion target covariance matrix and jamming covariance matrix are obtained. The method for obtaining the nonlinear motion target signal power vector in all distance gates comprises the following steps: a signal-to-jamming and noise ratio is obtained according to the nonlinear motion target covariance matrix and the jamming covariance matrix; an adaptive least mean square distortionless response beamformer is used to solve a maximum weight vector in the signal-to-jamming and noise ratio; and According to the maximum weight vector, the nonlinear moving target signal power on each range gate is obtained in sequence, and a nonlinear moving target power vector on all range gates is obtained.

6. The cognitive FDA-MIMO radar deception jamming rejection target detection method of claim 5, wherein, The construction process of the adaptive minimum mean square distortionless response beamformer includes: According to the estimated nonlinear target distance, angle and interference noise covariance matrix in the nonlinear moving target state, an adaptive minimum mean square distortionless response beamformer is constructed.

7. The cognitive FDA-MIMO radar deception jamming rejection target detection method of claim 1, wherein, The detection of the nonlinear moving target signal power vector on all range gates using the cascaded constant false alarm detector includes: The nonlinear moving target signal power vector on all range gates is detected using a preset first-stage constant false alarm detector, and the distance gate number and radial distance measurement of the nonlinear moving target are determined; a preset virtual steering vector, obtaining the signal power on the first angle resolution unit, and obtaining a signal power vector at the angle resolution unit; The nonlinear moving target angle detection information is obtained using a preset second-stage constant false alarm detector; The measurement value of the nonlinear moving target at the current time is obtained using the cascaded constant false alarm detector; The updated nonlinear moving target signal is obtained according to the measurement value of the nonlinear moving target at the current time.

8. The cognitive FDA-MIMO radar deception jamming rejection target detection method of claim 7, wherein, The process of obtaining the updated nonlinear moving target signal includes: The predicted measurement value of the sampling point set is solved; The measurement covariance matrix is obtained according to the predicted measurement value of the sampling point set and the measurement value of the nonlinear moving target at the current time; The state covariance matrix is obtained according to the predicted measurement value of the sampling point set and the measurement value of the nonlinear moving target at the current time; The nonlinear moving target state and the nonlinear moving target state covariance matrix at the current time are updated according to the measurement covariance matrix and the state covariance matrix.

Images