A method and device for suppressing multipath clutter of airborne radar

Abstract

The present invention relates to the field of signal processing technology, and provides a method and device for suppressing multipath clutter in an airborne radar. The method comprises: calculating the test statistic of each training sample in a sample set, selecting the training sample corresponding to the median of the test statistic as a reference sample; calculating the Mahalanobis distance between the reference sample and each training sample, using the Mahalanobis distance to perform non-uniform measurement on each training sample to screen out uniform training samples, and determining the unit sample to be detected where the multipath clutter is located; constructing the space-space domain clutter covariance matrix of the unit sample to be detected according to the unit sample to be detected, calculating the rank of the space-space domain clutter covariance matrix, thereby constructing a multi-stage Wiener filter and a blocking matrix, constructing an space-space domain adaptive weight vector according to the multi-stage Wiener filter and the blocking matrix; performing space-time joint processing according to the space-space domain adaptive weight vector to obtain an output signal. The present invention improves the target detection performance of the airborne early warning radar in a complex mountainous multipath clutter environment.

Description

Technical Field

[0001] The present invention relates to the technical field of signal processing, and in particular to a method and device for suppressing multipath clutter of an airborne radar. Background Art

[0002] Plateaus and mountainous areas are not only high in altitude, but also have a large number of mountains. Since airborne radars emit electromagnetic waves downward for detection, there will be more multipath clutter from mountainous areas compared to plains. At the same time, since the early warning aircraft is a moving platform, the platform height and platform movement will bring main clutter with Doppler components and curved clutter. Due to the true height and complex terrain of plateaus and mountainous areas, these clutters become stronger, which has a significant impact on radar detection, mainly including: (1) aliasing of real targets and multipath clutter; (2) increasing the clutter freedom; (3) increasing clutter non-uniformity; (4) the receiver is easy to saturate. The superposition of many influencing factors has caused multipath clutter to become one of the key technologies that urgently need to be solved for airborne radar downward detection.

[0003] The space-time adaptive processing (STAP) method can distinguish targets and clutter from both spatial and temporal dimensions, and is widely used in airborne radar clutter suppression. However, multipath clutter is mostly distributed near the mainlobe clutter in spatial domain, and there is a very obvious Doppler frequency broadening characteristic in the time domain. When the target falls into this frequency, we cannot suppress multipath clutter while retaining the target through the traditional space-time decoupling STAP method.

[0004] In view of this, overcoming the defects of the prior art is an urgent problem to be solved in the field of this technology. Summary of the invention

[0005] The technical problem to be solved by the present invention is to provide an airborne radar multipath clutter suppression method and device to improve the target detection performance of the airborne early warning radar in a complex mountainous multipath clutter environment.

[0006] The present invention adopts the following technical solution:

[0007] In a first aspect, the present invention provides an airborne radar multipath clutter suppression method, comprising:

[0008] Calculate the test statistic for each training sample in the sample set , select the training sample corresponding to the median of the test statistic as the reference sample ; Calculate reference sample The Mahalanobis distance to each training sample , using the Mahalanobis distance Perform non-uniform measurement on each training sample to screen out uniform training samples , and determine the unit sample to be detected where the multipath clutter is located ;

[0009] According to the unit sample to be detected , construct the unit sample to be detected The clutter covariance matrix of the empty space domain is , calculate the spatial clutter covariance matrix Rank ; According to the spatial domain clutter covariance matrix Rank , construct a multi-stage Wiener filter , and construct a multi-stage Wiener filter The blocking matrix , according to the multi-stage Wiener filter and the blocking matrix , construct the air-space domain adaptive weight vector ;

[0010] According to the spatial domain adaptive weight vector Perform space-time joint processing to obtain the output signal .

[0011] Preferably, the test statistic of each training sample is ;

[0012] The reference sample The Mahalanobis distance to each training sample ;in, represents the sample covariance matrix, , is the training sample, is the number of training samples in the sample set.

[0013] Preferably, the construction of the unit sample to be detected The clutter covariance matrix of the empty space domain is , calculate the spatial clutter covariance matrix Rank , including:

[0014] By the unit sample to be tested The slow time estimation obtains the spatial clutter covariance matrix ;in, K is the number of coherently accumulated pulses, for Slow time spatial domain sampling data;

[0015] Build m The transmission frequency and n The air-to-air steering vector of the receiving frequency is ,in, The spatial frequency is The launch airspace steering vector, The spatial frequency is The receiving spatial domain steering vector of

[0016] Calculate the m The transmission frequency and n The minimum variance spectrum of the received frequencies ;

[0017] According to the minimum variance spectrum , the spatial clutter covariance matrix is ​​calculated Rank ;in, F is an indicator function, , T is the decision threshold.

[0018] Preferably, the spatial domain clutter covariance matrix Rank , construct a multi-stage Wiener filter , including:

[0019] Construct the multi-stage Wiener filter ;in, ; , yes The blocking matrix of the components;

[0020] The multi-stage Wiener filter is constructed The blocking matrix , including:

[0021] Use the air-to-air guidance vector at the target location And Q equally spaced air-to-air guidance vectors at the left and right of the target position Constructing the blocking matrix of a multi-stage Wiener filter ;in, , M is the number of transmitting array elements of the airborne radar, N is the number of receiving array elements of the airborne radar, Q is the preset value, represents singular value decomposition, , , are the three matrices obtained by singular value decomposition, Represents the matrix of List to The columns form a new matrix.

[0022] Preferably, according to the multi-stage Wiener filter and the blocking matrix , construct the air-space domain adaptive weight vector , including:

[0023] The adaptive weight of the airspace domain after the first level of blocking is calculated as ;in, , , , , for Through the blocking matrix The processed signal, for Slow time spatial domain sampling data, , for go through The beamforming result is represents the mean;

[0024] Constructing the adaptive weight vector in the air-space domain ;in, is the air-to-air guidance vector of the target position.

[0025] Preferably, the spatial domain adaptive weight vector Perform space-time joint processing to obtain the output signal , including:

[0026] According to the spatial domain adaptive weight vector , construct the space-time dimension reduction matrix ;

[0027] Using the space-time dimension reduction matrix Samples of units to be tested Perform dimensionality reduction processing to obtain the first dimensionality reduction sample , using the space-time dimension reduction matrix For uniform training samples Perform dimensionality reduction processing to obtain the second dimensionality reduction sample ;

[0028] According to the first dimension reduction sample and the second dimension reduction sample , construct space-time three-dimensional weight vector , using the space-time three-dimensional weight vector For the first dimension reduction sample Processing to get the output signal .

[0029] Preferably, the spatial domain adaptive weight vector , construct the space-time dimension reduction matrix , including:

[0030] Constructing space-time dimension reduction matrix ;in, Indicates k The dimension corresponding to the Doppler channel K × K' The Doppler filter matrix, , , and represents the Doppler filter weight vector corresponding to the k-1th, kth and k+1th Doppler channels, Representation Dimension MN × N' The beamforming matrix is , , , ; is the target location, and are the two closest equally spaced positions on the left and right of the target position, It means that the target orientation vector is obtained as The MWF weight vector.

[0031] Preferably, the space-time dimension reduction matrix is ​​used Samples of units to be tested Perform dimensionality reduction processing to obtain the first dimensionality reduction sample , using the space-time dimension reduction matrix For uniform training samples Perform dimensionality reduction processing to obtain the second dimensionality reduction sample ; According to the first dimension reduction sample and the second dimension reduction sample , construct space-time three-dimensional weight vector , using the space-time three-dimensional weight vector For the first dimension reduction sample Processing to get the output signal , including:

[0032] Calculate the first dimension reduction sample ; Calculate the second dimension reduction sample , For uniform training samples the number of

[0033] According to the first dimension reduction sample and the second dimension reduction sample , determine the k The space-time three-dimensional covariance matrix corresponding to the range-Doppler unit to be detected ;

[0034] Constructing space-time 3D weight vector ;in, For the The reduced-dimensional space-time steering vector of the Doppler channel, ; is the three-dimensional guidance vector in the empty space, , is the time-domain steering vector, is the launch airspace steering vector, To receive the spatial steering vector;

[0035] Using the space-time three-dimensional weight vector For the first dimension reduction sample Processing to get the output signal .

[0036] In a second aspect, the present invention further provides an airborne radar multipath clutter suppression device, which is used to implement the airborne radar multipath clutter suppression method described in the first aspect, and the device includes:

[0037] At least one processor; and a memory in communication with the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the processor to execute the airborne radar multipath clutter suppression method described in the first aspect.

[0038] In a third aspect, the present invention further provides a non-volatile computer storage medium, wherein the computer storage medium stores computer executable instructions, and the computer executable instructions are executed by one or more processors to complete the method described in the first aspect.

[0039] In a fourth aspect, a chip is provided, comprising: a processor and an interface, for calling and running a computer program stored in the memory from a memory, and executing the method of the first aspect.

[0040] According to a fifth aspect, a computer program product comprising instructions is provided. When the instructions are executed on a computer or a processor, the computer or the processor executes the method according to the first aspect.

[0041] The present invention firstly uses Mahalanobis distance to measure the non-uniformity of distance dimension training samples, and preliminarily determines the distance gate where the multipath clutter is located (that is, determines the sample unit to be detected where the multipath clutter is located); secondly, the rank of the multipath clutter covariance matrix is ​​estimated by using the minimum variance spectrum to determine the order of the multi-stage Wiener filter; thirdly, a filtering weight vector is formed in the air-space domain by constructing the blocking matrix of the multi-stage Wiener filter; finally, multipath clutter is suppressed by space-time joint processing, thereby improving the target detection performance of airborne early warning radar in a complex mountainous multipath clutter environment. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the embodiments of the present invention. Obviously, the drawings described below are only some embodiments of the present invention, and for ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0043] Figure 1 It is a flow chart of a method for suppressing multipath clutter of an airborne radar provided by an embodiment of the present invention;

[0044] Figure 2 It is a flow chart of a method for suppressing multipath clutter of an airborne radar provided by an embodiment of the present invention;

[0045] Figure 3 It is a flow chart of a method for suppressing multipath clutter of an airborne radar provided by an embodiment of the present invention;

[0046] Figure 4 It is a flow chart of a method for suppressing multipath clutter of an airborne radar provided by an embodiment of the present invention;

[0047] Figure 5 It is a flow chart of a method for suppressing multipath clutter of an airborne radar provided by an embodiment of the present invention;

[0048] Figure 6 is a schematic diagram of an airborne radar multipath clutter suppression method provided by an embodiment of the present invention;

[0049] Figure 7 1 is a schematic diagram of the architecture of an airborne radar multipath clutter suppression device provided by an embodiment of the present invention;

[0050] Figure 8 The present invention is a schematic diagram of the architecture of an airborne radar multipath clutter suppression device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0051] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0052] Unless the context requires otherwise, throughout the specification and claims, the term "including" is to be interpreted as open inclusion, that is, "including, but not limited to". In the description of the specification, the terms "one embodiment", "some embodiments", "exemplary embodiments", "examples", "specific examples" or "some examples" and the like are intended to indicate that specific features, structures, materials or characteristics associated with the embodiment or example are included in at least one embodiment or example of the present disclosure. The schematic representation of the above terms does not necessarily refer to the same embodiment or example. In addition, the specific features, structures, materials or characteristics may be included in any one or more embodiments or examples in any appropriate manner, that is, although they may be carried in the embodiments or examples of the above terms due to reasons such as the order and position of appearance, it is not limited to that they can be carried in combination by one embodiment or example.

[0053] In the description of the present invention, the terms "first" and "second" are used only for descriptive purposes, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of indicated technical features. Thus, the features defined as "first" and "second" may explicitly or implicitly include one or more of the features. In the description of the embodiments of the present disclosure, unless otherwise specified, the meaning of "multiple" is two or more. In addition, for example, the same type of nouns may be described as two independent individuals by adding "A" and "B" at the end. In this case, the corresponding features defined as "A" and "B" are only used to distinguish the same type of individuals for description purposes, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of indicated technical features.

[0054] In the description of the present invention, the expression "A and / or B" (where A and B are used to formally represent specific characteristic contents) is involved, and the corresponding expressions include the following three combinations: only A, only B, and a combination of A and B.

[0055] As used herein, "about," "substantially," or "approximately" includes the stated value and an average value that is within an acceptable range of deviation from the particular value as determined by one of ordinary skill in the art taking into account the measurements in question and the errors associated with the measurement of the particular quantity (i.e., the limitations of the measurement system).

[0056] In addition, the technical features involved in the various embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0057] Embodiment 1:

[0058] To address the problem of multipath clutter suppression, Deng et al. proposed a novel cascade clutter suppression method. This method introduces a cost function with sparse constraints, and then uses an l1-based recursive least squares algorithm to calculate the optimization problem and the multipath clutter suppression weight vector. However, the method has the problem of insufficient prior knowledge, resulting in performance degradation and excessive computation. Krolik et al. used the sidelobe cancellation structure to form a partial adaptive multiple-input multiple-output (MIMO) beamforming weight vector to achieve multipath clutter suppression. This method has a significant effect in suppressing multipath clutter during vehicle-mounted radar target detection, but this method is not suitable for airborne platforms.

[0059] In actual engineering, when the airborne radar is working downward, in addition to the inevitable near- and long-range non-stationary and non-uniform clutter, there may also be a mountain multipath clutter environment, and the clutter non-uniformity is significantly increased. The existence of multipath clutter will seriously affect the target detection performance of the airborne early warning radar, and the existing clutter suppression methods cannot effectively suppress multipath clutter. Therefore, the detection of moving targets in a complex mountain multipath clutter environment has become one of the core key technical problems that the current airborne radar urgently needs to solve. In order to solve this problem, Embodiment 1 of the present invention provides an airborne radar multipath clutter suppression method, such as Figure 1 As shown, including:

[0060] In step 201, the test statistic of each training sample in the sample set is calculated , select the training sample corresponding to the median of the test statistic as the reference sample ; Calculate reference sample The Mahalanobis distance to each training sample , using the Mahalanobis distance Perform non-uniform measurement on each training sample to screen out uniform training samples , and determine the unit sample to be detected where the multipath clutter is located ; Specifically: each sample in the sample training set is divided into The values ​​are sorted from small to large, according to The value of selects the reference sample, similar to the median filter method, and selects The sample corresponding to the median is used as the reference sample for subsequent non-uniform clutter data processing. The sample training set consists of L The determination of the unit sample to be detected where the multipath clutter is located It can be: When the corresponding sample When it is greater than the preset value, the sample is considered to be a unit sample to be detected. The preset value is obtained by those skilled in the art based on empirical analysis and can be understood as: When the Mahalanobis distance is large, the sample contains not only the target and direct wave clutter, but also multipath clutter. At this time, it is necessary to decouple the target and multipath clutter in the air domain. Also known as the generalized inner product. Using the Mahalanobis distance Perform non-uniform measurement on each training sample to screen out uniform training samples Specifically: By comparing all training samples in the scene with the reference samples The Mahalanobis distance , and using the Mahalanobis distance To perform non-uniform measurement on all training samples.

[0061] In step 202, according to the unit sample to be detected , construct the unit sample to be detected The clutter covariance matrix of the empty space domain is , calculate the spatial clutter covariance matrix Rank ; According to the spatial domain clutter covariance matrix Rank , construct a multi-stage Wiener filter , and construct a multi-stage Wiener filter The blocking matrix , according to the multi-stage Wiener filter and the blocking matrix , construct the air-space domain adaptive weight vector .

[0062] In step 203, according to the spatial domain adaptive weight vector Perform space-time joint processing to obtain the output signal .

[0063] This embodiment first uses the Mahalanobis distance to measure the non-uniformity of the distance dimension training samples, and preliminarily determines the range gate where the multipath clutter is located (that is, determines the sample unit to be detected where the multipath clutter is located); secondly, the rank of the multipath clutter covariance matrix is ​​estimated by the minimum variance spectrum to determine the order of the multi-stage Wiener filter; thirdly, a blocking matrix of the multi-stage Wiener filter is constructed to form a filter weight vector in the air-space domain; finally, multipath clutter is suppressed by space-time joint processing, thereby improving the target detection performance of the airborne early warning radar in a complex mountainous multipath clutter environment.

[0064] In a specific application scenario, the test statistic of each training sample is The reference sample The Mahalanobis distance to each training sample ;in, represents the sample covariance matrix, It is estimated from the samples in the initial training sample set. , is the training sample, is the number of training samples in the sample set.

[0065] In some embodiments, the construction of the unit sample to be detected The clutter covariance matrix of the empty space domain is , calculate the spatial clutter covariance matrix Rank ,like Figure 2 As shown, specifically including:

[0066] In step 301, the unit sample to be detected The slow time estimation obtains the spatial clutter covariance matrix ;in, K is the number of coherently accumulated pulses, for Slow time spatial domain sampling data.

[0067] In step 302, since the number of sampling points in the transmitting frequency domain and the receiving frequency domain is consistent with the number of transmitting and receiving array elements, the first m The transmission frequency and n The air-to-air steering vector of the receiving frequency is , calculate the m The transmission frequency and n The minimum variance spectrum of the received frequencies ; Minimum variance spectrum The real power distribution of clutter plus noise in the two-dimensional plane of the air space domain is characterized.

[0068] In step 303, according to the minimum variance spectrum , the spatial clutter covariance matrix is ​​calculated Rank ;in, F is an indicator function, , T is the decision threshold.

[0069] In an optional implementation, the spatial clutter covariance matrix Rank , construct a multi-stage Wiener filter , specifically comprising: constructing the multi-stage Wiener filter ; This multi-stage Wiener filter is also called a reduced rank matrix, and its dimension is (MN-1)×D , where, according to adaptive filtering theory, D The value of needs to satisfy: ; , The dimension of (MN-1)×1 , yes The blocking matrix of the components.

[0070] In order to ensure high gain at the target of the airspace filter and avoid distortion of the direct wave clutter near the target, the position of the blocking matrix is ​​selected at the target and Q equally spaced positions on the left and right of the target, that is, the multi-stage Wiener filter is constructed. The blocking matrix , specifically: using the air-to-air guidance vector at the target position And Q equally spaced air-to-air guidance vectors at the left and right of the target position Constructing the blocking matrix of a multi-stage Wiener filter ;in, , M is the number of transmitting array elements of the airborne radar, N is the number of receiving array elements of the airborne radar, and Q is a preset value obtained by those skilled in the art based on empirical analysis. represents singular value decomposition, , , are the three matrices obtained by singular value decomposition, Represents the matrix of List to The columns form a new matrix, where the equally spaced positions are actually positions with equally spaced azimuth angles adjacent to the target position on the left and right. For example, in an actual application scenario, Q=10, then the Q equally spaced positions on the left and right sides of the target are 5 adjacent positions with an azimuth angle interval of 1° on the left and right sides of the target position.

[0071] According to the multi-stage Wiener filter and the blocking matrix , construct the air-space domain adaptive weight vector , specifically including: calculating the air-space adaptive weight after the first level of blocking is ;in, , , , , for Through the blocking matrix The processed signal, for Slow time spatial domain sampling data, , for go through The beamforming result is Represents the mean; constructs the air-space domain adaptive weight vector ;in, is the air-to-air guidance vector of the target position.

[0072] Considering that the space-time dimension reduction filtering process is separable, that is, the spatial domain beamforming and Doppler filtering can be cascaded, the spatial domain adaptive weight vector Perform space-time joint processing to obtain the output signal ,like Figure 3 As shown, specifically including:

[0073] In step 401, according to the spatial domain adaptive weight vector , construct the space-time dimension reduction matrix .

[0074] In step 402, the space-time dimension reduction matrix is ​​used Samples of units to be tested Perform dimensionality reduction processing to obtain the first dimensionality reduction sample , using the space-time dimension reduction matrix For uniform training samples Perform dimensionality reduction processing to obtain the second dimensionality reduction sample .

[0075] In step 403, according to the first dimension reduction sample and the second dimension reduction sample , construct space-time three-dimensional weight vector , using the space-time three-dimensional weight vector For the first dimension reduction sample Processing to get the output signal .

[0076] The spatial domain adaptive weight vector , construct the space-time dimension reduction matrix , specifically including: constructing a space-time dimension reduction matrix ;in, Indicates k The dimension corresponding to the Doppler channel K × K' The Doppler filter matrix is ​​assumed to be K' =3, N' =3, The channel corresponds to K ×1-dimensional DFT weight vector, , , and represents the Doppler filter weight vector corresponding to the k-1th, kth and k+1th Doppler channels, Representation Dimension MN × N' The beamforming matrix is , , , ; is the target location, and are the two closest equally spaced positions on the left and right of the target position, It means that the target orientation vector is obtained as MWF weight vector. In actual use, and The positions on the left and right sides of the target are 1° apart from the target position in azimuth. , and The corresponding values ​​are the target and the two closest equally spaced positions on the left and right of the target. MN×1 dimensional MWF weight vector.

[0077] The use of the space-time dimension reduction matrix Samples of units to be tested Perform dimensionality reduction processing to obtain the first dimensionality reduction sample , using the space-time dimension reduction matrix For uniform training samples Perform dimensionality reduction processing to obtain the second dimensionality reduction sample ; According to the first dimension reduction sample and the second dimension reduction sample , construct space-time three-dimensional weight vector , using the space-time three-dimensional weight vector For the first dimension reduction sample Processing to get the output signal ,like Figure 4 As shown, specifically including:

[0078] In step 501, the first dimension reduction sample is calculated ; Calculate the second dimension reduction sample , For uniform training samples The number of units to be tested Also known as the unit data to be detected, the uniform training sample Also called training unit data.

[0079] In step 502, according to the first dimension reduction sample and the second dimension reduction sample , determine the k The space-time three-dimensional covariance matrix corresponding to the range-Doppler unit to be detected .

[0080] In step 503, a space-time three-dimensional weight vector is constructed ;in, For the The reduced-dimensional space-time steering vector of the Doppler channel, ; is the three-dimensional guidance vector in the empty space, , is the time-domain steering vector, is the launch airspace steering vector, is the receiving spatial steering vector.

[0081] In step 504, the space-time three-dimensional weight vector For the first dimension reduction sample Processing to get the output signal .

[0082] This embodiment uses the receiving and transmitting spatial frequency of the airborne radar to distinguish between targets, multipath clutter and direct wave clutter. On the one hand, the accurate positioning of multipath clutter in the radar echo range Doppler map is achieved through non-uniform measurement. On the other hand, the airborne radar air-space domain beamforming is used to suppress multipath clutter, which is of great significance for improving the target detection performance of airborne early warning radar in complex mountainous multipath clutter environment.

[0083] Embodiment 2:

[0084] The present invention is based on the method described in Example 1, combined with specific application scenarios, and uses technical descriptions in related scenarios to illustrate the implementation process of the present invention in characteristic scenarios.

[0085] This embodiment uses Figure 6 Taking the airborne radar geometric configuration scenario shown in FIG. 1 as an example, the airborne radar multipath clutter suppression method provided in this embodiment is as follows: Figure 5 As shown, the following steps are included:

[0086] In step 601, the generalized inner product of each training sample in the sample set is calculated, and a reference sample is selected based on the generalized inner product. Then, the Mahalanobis distance between the training sample and the reference sample is calculated respectively, and the Mahalanobis distance is used to perform non-uniform measurement on all samples.

[0087] In step 602, the rank of the spatial domain clutter covariance matrix of the unit sample to be detected is calculated, and the order of the multi-stage Wiener filter is determined according to the size of the matrix rank.

[0088] In step 603, according to the position of the target in the air-space domain, the blocking matrices corresponding to the multiple transmit and receive beams are flexibly selected to ensure the target gain, and then the air-space domain MWF weight vectors of each range gate are constructed.

[0089] In step 604, adjacent 3-channel space-time joint dimension reduction STAP processing is used to further suppress the residual clutter and retain the target signal. The number of transmitting array elements of the airborne radar is set to M, the number of receiving array elements is set to N, and the decomposition order of the MWF method is set to D.

[0090] The step 601 specifically includes:

[0091] Set the number of transmitting array elements of the airborne radar to M , the number of receiving array elements is N , the decomposition order of the MWF method is D .

[0092] Assume that the initial training sample set is composed of L training samples, then l The GIP test statistic corresponding to the sample can be expressed as:

[0093] (1)

[0094] in represents the sample covariance matrix. Usually It is estimated from the samples in the initial training sample set, that is:

[0095] (2)

[0096] The sample training set according to The values ​​are sorted from small to large, according to The value of selects the reference sample, similar to the median filter method, and selects The sample corresponding to the median is used as the reference sample for subsequent non-uniform clutter data processing. After selecting the reference samples, calculate the reference samples separately and all samples The Mahalanobis distance between them is:

[0097] (3)

[0098] The step 602 specifically includes:

[0099] When the unit sample to be tested When the Mahalanobis distance of is large, the sample contains not only the target and direct wave clutter, but also multipath clutter. In this case, it is necessary to decouple the target and multipath clutter in the air domain. First, we need to estimate The clutter covariance matrix of the empty space domain is , which can be estimated from the slow time of the unit sample to be detected ,Right now:

[0100] (4)

[0101] in, K is the number of coherently accumulated pulses, for The slow time spatial domain sampling data. Since the number of sampling points in the transmitting frequency domain and the receiving frequency domain is consistent with the number of transmitting and receiving array elements, the m The transmission frequency and n The air-to-air steering vector of a receiving frequency can be expressed as:

[0102] (5)

[0103] Since the minimum variance spectrum The real power distribution of clutter and noise in the two-dimensional plane of the air domain is characterized, so the power of the receiving and transmitting frequency can be calculated. for:

[0104] (6)

[0105] According to the two-dimensional plane of the empty space MN Pointed Estimable Rank for:

[0106] (7)

[0107] in F is an indicator function, T is the decision threshold, which is expressed as:

[0108] (8)

[0109] The step 603 specifically includes:

[0110] The mathematical expression of the reduced rank matrix of the MWF method is:

[0111] (9)

[0112] Its dimension is ( MN -1)× D , according to the adaptive filtering theory, D The value of needs to satisfy: ,in:

[0113] (10)

[0114] The dimension of is ( MN -1)×1, yes The blocking matrix of the components.

[0115] Then the air-space domain adaptive weight of the MWF method after the first level of blocking can be expressed as:

[0116] (11)

[0117] in:

[0118] (12)

[0119] (13)

[0120] in, , , for Through the blocking matrix The processed signal, , for go through The beamforming result is Represents the mean.

[0121] Therefore, the total D The adaptive weights in the air-space domain of the order MWF method are:

[0122] (14)

[0123] In order to ensure high gain at the target of the airspace filter and avoid distortion of the direct wave clutter near the target, the position of the blocking matrix in equation (9) is selected at the target and the 5 adjacent azimuths of the target with an interval of 1°. A method of calculating the blocking matrix using singular value decomposition for:

[0124] (15)

[0125] in, The target and the 5 adjacent positions on the left and right sides of the target with an azimuth interval of 1° are included. Q +1 position of air-to-air guidance vector, Q =10, represents singular value decomposition, , , are the three matrices obtained by singular value decomposition, Represents the matrix of List to The columns form a new matrix.

[0126] The step 604 specifically includes:

[0127] Considering that the space-time dimensionality reduction filtering process is separable, that is, the spatial domain beamforming and Doppler filtering can be cascaded. Then the space-time dimensionality reduction matrix It can be expressed as:

[0128] (16)

[0129] in Indicates k The dimension corresponding to the Doppler channel K × K' The Doppler filter matrix, Representation Dimension MN × N' The beamforming matrix is ​​assumed to be K ' =3, N ' =3, The channel corresponds to K ×1-dimensional DFT weight vector, that is , and represents the Doppler filter weight vector corresponding to the k-1th, kth and k+1th Doppler channels, then It can be expressed as:

[0130] (17)

[0131] , and Indicates the position of the target and the two adjacent azimuth angles on the left and right of the target with an interval of 1° MN ×1-dimensional MWF weight vector, , , , It means that the target orientation vector is obtained as The MWF weight vector is It can be expressed as:

[0132] (18)

[0133] The unit data to be detected and the training unit data after dimensionality reduction are:

[0134] (19)

[0135] (20)

[0136] in L s is the number of training samples. It should be noted that is a uniform training sample selected from the initial sample set by non-uniform measurement. k The space-time three-dimensional covariance matrix corresponding to the range-Doppler unit to be detected is:

[0137] (twenty one)

[0138] The space-time three-dimensional weight vector of the range-Doppler unit to be detected is:

[0139] (twenty two)

[0140] in, For the The reduced-dimensional space-time steering vector of the Doppler channel, is the three-dimensional guidance vector in the empty space.

[0141] The processed output signal is:

[0142] (twenty three)

[0143] Embodiment 3:

[0144] like Figure 7 As shown, this embodiment also provides an airborne radar multipath clutter suppression device, including a non-uniformity measurement module, a clutter rank estimation module, an air-space domain beamforming module and a space-time joint processing module.

[0145] The non-uniformity measurement module calculates the generalized inner product of each training sample in the sample set, and selects a reference sample based on this, then calculates the Mahalanobis distance between the training sample and the reference sample, and uses the Mahalanobis distance to perform non-uniformity measurement on all samples.

[0146] The clutter rank estimation module calculates the rank of the spatial domain clutter covariance matrix of the unit sample to be detected, and determines the order of the multi-stage Wiener filter according to the size of the matrix rank.

[0147] The air-space domain beamforming module flexibly selects blocking matrices corresponding to multiple transceiver beams according to the position of the target in the air-space domain, thereby ensuring the target gain, and then constructs the air-space domain MWF weight vector of each range gate.

[0148] The space-time joint processing module adopts adjacent 3-channel space-time joint dimension reduction STAP processing to further suppress residual clutter and retain target signals.

[0149] like Figure 8 FIG. 1 is another schematic diagram of the architecture of an airborne radar multipath clutter suppression device according to an embodiment of the present invention. The airborne radar multipath clutter suppression device according to this embodiment includes one or more processors 21 and a memory 22. Figure 8 A processor 21 is taken as an example.

[0150] The processor 21 and the memory 22 may be connected via a bus or other means. Figure 8 The example of connecting through bus is taken in the following.

[0151] The memory 22 is a non-volatile computer-readable storage medium that can be used to store non-volatile software programs and non-volatile computer executable programs, such as the airborne radar multipath clutter suppression method in Example 1. The processor 21 executes the airborne radar multipath clutter suppression method by running the non-volatile software programs and instructions stored in the memory 22.

[0152] The memory 22 may include a high-speed random access memory, and may also include a non-volatile memory, such as at least one disk storage device, a flash memory device, or other non-volatile solid-state storage devices. In some embodiments, the memory 22 may optionally include a memory remotely arranged relative to the processor 21, and these remote memories may be connected to the processor 21 via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0153] The program instructions / modules are stored in the memory 22, and when executed by the one or more processors 21, the airborne radar multipath clutter suppression method in the above-mentioned embodiment 1 is executed.

[0154] It is worth noting that the information interaction, execution process, etc. between the modules and units within the above-mentioned devices and systems are based on the same concept as the processing method embodiment of the present invention. The specific contents can be found in the description of the method embodiment of the present invention and will not be repeated here.

[0155] A person skilled in the art may understand that all or part of the steps in the various methods of the embodiments may be completed by instructing related hardware through a program, and the program may be stored in a computer-readable storage medium, and the storage medium may include: a read-only memory (ROM), a random access memory (RAM), a disk or an optical disk, etc.

[0156] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for suppressing multipath clutter in an airborne radar, characterized in that: include: Calculate the test statistic for each training sample in the sample set , select the training sample corresponding to the median of the test statistic as the reference sample ; Calculate reference sample The Mahalanobis distance to each training sample , using the Mahalanobis distance Perform non-uniform measurement on each training sample to screen out uniform training samples , and determine the unit sample to be detected where the multipath clutter is located ; According to the unit sample to be detected , construct the unit sample to be detected The clutter covariance matrix of the empty space domain is , calculate the spatial clutter covariance matrix Rank ; According to the spatial domain clutter covariance matrix Rank , construct a multi-stage Wiener filter , and construct a multi-stage Wiener filter The blocking matrix , according to the multi-stage Wiener filter and the blocking matrix , construct the air-space domain adaptive weight vector ; According to the spatial domain adaptive weight vector Perform space-time joint processing to obtain the output signal .

2. The airborne radar multipath clutter suppression method according to claim 1, characterized in that: The test statistic of each training sample ; The reference sample The Mahalanobis distance to each training sample ;in, represents the sample covariance matrix, , is the training sample, is the number of training samples in the sample set.

3. The airborne radar multipath clutter suppression method according to claim 1, characterized in that: The construction of the unit sample to be detected The clutter covariance matrix of the empty space domain is , calculate the spatial clutter covariance matrix Rank , specifically including: By the unit sample to be tested The slow time estimation obtains the spatial clutter covariance matrix ;in, K is the number of coherently accumulated pulses, for Slow time spatial domain sampling data; Build m The transmission frequency and n The air-to-air steering vector of the receiving frequency is ,in, The spatial frequency is The launch airspace steering vector, The spatial frequency is The receiving spatial domain steering vector of Calculate the m The transmission frequency and n The minimum variance spectrum of the received frequencies ; According to the minimum variance spectrum , the spatial clutter covariance matrix is ​​calculated Rank ;in, F is an indicator function, , T is the decision threshold.

4. The airborne radar multipath clutter suppression method according to claim 1, characterized in that: The spatial clutter covariance matrix according to the Rank , construct a multi-stage Wiener filter , specifically including: Construct the multi-stage Wiener filter ;in, ; , yes The blocking matrix of the components; The multi-stage Wiener filter is constructed The blocking matrix , specifically including: Use the air-to-air guidance vector at the target location And Q equally spaced air-to-air guidance vectors at the left and right of the target position Constructing the blocking matrix of a multi-stage Wiener filter ;in, , M is the number of transmitting array elements of the airborne radar, N is the number of receiving array elements of the airborne radar, Q is the preset value, represents singular value decomposition, , , are the three matrices obtained by singular value decomposition, Represents the matrix of List to The columns form a new matrix.

5. The airborne radar multipath clutter suppression method according to claim 1, characterized in that: According to the multi-stage Wiener filter and the blocking matrix , construct the air-space domain adaptive weight vector , specifically including: The adaptive weight of the airspace domain after the first level of blocking is calculated as ;in, , , , , for Through the blocking matrix The processed signal, for Slow time spatial domain sampling data, , for go through The beamforming result is represents the mean; Constructing the adaptive weight vector in the air-space domain ;in, is the air-to-air guidance vector of the target position.

6. The method for suppressing multipath clutter of an airborne radar according to claim 1, characterized in that: The spatial domain adaptive weight vector Perform space-time joint processing to obtain the output signal , specifically including: According to the spatial domain adaptive weight vector , construct the space-time dimension reduction matrix ; Using the space-time dimension reduction matrix Samples of units to be tested Perform dimensionality reduction processing to obtain the first dimensionality reduction sample , using the space-time dimension reduction matrix For uniform training samples Perform dimensionality reduction processing to obtain the second dimensionality reduction sample ; According to the first dimension reduction sample and the second dimension reduction sample , construct space-time three-dimensional weight vector , using the space-time three-dimensional weight vector For the first dimension reduction sample Processing to get the output signal .

7. The method for suppressing multipath clutter of airborne radar according to claim 6, characterized in that: The spatial domain adaptive weight vector , construct the space-time dimension reduction matrix , specifically including: Constructing space-time dimension reduction matrix ;in, Indicates k The dimension corresponding to the Doppler channel K × K' The Doppler filter matrix, , , and represents the Doppler filter weight vector corresponding to the k-1th, kth and k+1th Doppler channels, Representation Dimension MN × N' The beamforming matrix is , , , ; is the target location, and are the two closest equally spaced positions on the left and right of the target position, It means that the target orientation vector is obtained as The MWF weight vector.

8. The method for suppressing multipath clutter of airborne radar according to claim 6, characterized in that: The use of the space-time dimension reduction matrix Samples of units to be tested Perform dimensionality reduction processing to obtain the first dimensionality reduction sample , using the space-time dimension reduction matrix For uniform training samples Perform dimensionality reduction processing to obtain the second dimensionality reduction sample ; According to the first dimension reduction sample and the second dimension reduction sample , construct space-time three-dimensional weight vector , using the space-time three-dimensional weight vector For the first dimension reduction sample Processing to get the output signal , specifically including: Calculate the first dimension reduction sample ; Calculate the second dimension reduction sample , For uniform training samples the number of According to the first dimension reduction sample and the second dimension reduction sample , determine the k The space-time three-dimensional covariance matrix corresponding to the range-Doppler unit to be detected ; Constructing space-time 3D weight vector ;in, For the The reduced-dimensional space-time steering vector of the Doppler channel, ; is the three-dimensional guidance vector in the empty space, , is the time-domain steering vector, is the launch airspace steering vector, To receive the spatial steering vector; Using the space-time three-dimensional weight vector For the first dimension reduction sample Processing to get the output signal .

9. A non-volatile computer storage medium, characterized in that: The computer storage medium stores computer executable instructions, which are executed by one or more processors to complete the airborne radar multipath clutter suppression method according to any one of claims 1 to 8.

10. An airborne radar multipath clutter suppression device, characterized in that: include: at least one processor; And, a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the processor to execute the airborne radar multipath clutter suppression method according to any one of claims 1-8.

Images