Multi-channel main lobe interference separation and speed information extraction method based on MIMO radar

Abstract

The invention discloses a multi-channel main lobe interference separation and speed information extraction method and device based on an MIMO radar, and belongs to the technical field of radar signal processing. The method comprises the following steps: in each pulse repetition period, acquiring echo signals of multiple channels of the MIMO radar, and performing pulse compression processing to obtain echo signals after pulse compression; processing the echo signal after pulse compression by using an MDL algorithm, and estimating the number of signal sources; based on the estimated number of the signal sources, separating the echo signals after pulse compression by using a JADE blind source separation algorithm to obtain separated signals; and carrying out target identification on the separated signals obtained in all pulse repetition periods by using a peak sidelobe ratio technology, and carrying out coherent accumulation on the identified target signals so as to extract the speed information of the target. According to the method, the multi-channel characteristic of the MIMO radar system is utilized to effectively solve the problem of target speed information loss after separation.

Description

Technical Field

[0001] The invention belongs to the technical field of radar signal processing, and in particular relates to a multi-channel main lobe interference separation and velocity information extraction method based on MIMO radar. Background Art

[0002] With the widespread application of radar systems and the continuous development of external electronic interference methods, the radar is increasingly affected by external interference signals in the process of detecting targets. Traditional anti-interference methods usually rely on technologies such as low sidelobe blanking, sidelobe cancellation, frequency agility, and spatial beamforming. Although these methods can reduce interference to a certain extent, their effects are limited when facing complex multi-source interference or strong mainlobe interference. MIMO (Multiple Input Multiple Output) radar technology realizes joint signal processing in space, time, and frequency dimensions through the combination of multiple transmitting antennas and receiving antennas, greatly improving the resolution, detection performance, and anti-interference ability of the radar system. In a complex electromagnetic environment, one of the main challenges facing radar systems is the mainlobe interference problem, that is, the interference signal directly enters the main beam direction of the radar antenna, resulting in the aliasing of the target signal and the interference signal. Because this interference is in the same direction as the target signal, traditional sidelobe suppression, beam steering, and other methods are difficult to effectively deal with.

[0003] In MIMO radar systems, multi-channel data can be processed in two dimensions of space and time to separate signal sources from different directions and paths. Combined with spatial diversity and signal independence, it can more effectively distinguish target signals from interference signals. This provides good conditions for the use of advanced signal processing technologies such as blind source separation (BSS). BSS analyzes the statistical characteristics of signals without the need for a priori interference models and can adaptively separate multiple signal sources from mixed observation signals. With the help of multi-channel data from MIMO radar, BSS technology can effectively separate main lobe interference from target echo signals, improving the system's anti-interference capability, especially in complex electromagnetic environments.

[0004] Current single-channel radar interference separation methods mainly rely on the speed difference between the target and the interference, and use the difference in signal Doppler frequency shift to achieve blind source separation of the target and interference signals. This type of method can effectively separate interference and target signals to a certain extent, but there is still a key problem: the separated target signal cannot be processed coherently, that is, its phase information is lost during the separation process. This lack of phase information directly affects the coherence of the target signal, which in turn leads to errors or even failure in the extraction and estimation of target velocity information.

[0005] At present, there is no mature solution to the problem of velocity information loss in the separated target signal. This problem stems from the fact that existing methods cannot guarantee the phase consistency of the target signal while suppressing interference, especially in complex interference environments. Although existing research has made some progress in improving the accuracy and robustness of signal separation, maintaining the complete phase information of the target signal during the separation process remains an unresolved technical challenge. Summary of the invention

[0006] In order to solve the above problems existing in the prior art, the present invention provides a multi-channel main lobe interference separation and velocity information extraction method based on MIMO radar. The technical problem to be solved by the present invention is achieved by the following technical solutions:

[0007] In a first aspect, the present invention proposes a multi-channel main lobe interference separation and velocity information extraction method based on MIMO radar, comprising:

[0008] In each pulse repetition period, the multi-channel echo signal of the MIMO radar is obtained, and pulse compression processing is performed to obtain the echo signal after pulse compression;

[0009] The echo signal after pulse compression is processed using the MDL (Minimum Description Length) algorithm to estimate the number of signal sources.

[0010] Based on the estimated number of signal sources, the JADE (Joint Approximate Diagonalization of Eigenmatrices) blind source separation algorithm is used to separate the echo signal after pulse compression to obtain the separated signal;

[0011] For the separated signals obtained from all pulse repetition periods, the peak-to-sidelobe ratio technique is used for target identification, and the identified target signals are coherently integrated to extract the target's velocity information.

[0012] In a second aspect, the present invention proposes a multi-channel main lobe interference separation and velocity information extraction device based on MIMO radar, which is used to implement the multi-channel main lobe interference separation and velocity information extraction method based on MIMO radar proposed in the first aspect of the present invention, including:

[0013] The pulse compression processing module is used to obtain the echo signal of the multi-channel of the MIMO radar in each pulse repetition period, and perform pulse compression processing to obtain the echo signal after pulse compression;

[0014] A signal source estimation module is used to process the echo signal after pulse compression using the MDL algorithm to estimate the number of signal sources;

[0015] A signal separation module is used to separate the echo signal after pulse compression by using the JADE blind source separation algorithm based on the estimated number of signal sources to obtain a separated signal;

[0016] The signal extraction module is used to identify the target using the peak-to-sidelobe ratio technology for the separated signals obtained from all pulse repetition periods, and to perform coherent accumulation on the identified target signals to extract the target's speed information.

[0017] Beneficial effects of the present invention:

[0018] The multi-channel mainlobe interference separation and speed information extraction method based on MIMO radar provided by the present invention first obtains the echo signal of the multi-channel of the MIMO radar in each pulse repetition period and performs pulse compression processing; then uses the MDL algorithm to process the echo signal after pulse compression to estimate the number of signal sources; then uses the JADE blind source separation algorithm to separate the echo signal after pulse compression; finally uses the peak sidelobe ratio technology to identify the target signal of the separated signal, and repeats this operation in each pulse repetition period, and performs coherent accumulation on the identified target signal to extract the speed information of the target. The method uses the multi-channel characteristics of the MIMO radar system to effectively solve the problem of target speed information loss after separation.

[0019] The present invention will be further described in detail below with reference to the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 This is a schematic diagram of intermittent sampling direct forwarding interference provided by an embodiment of the present invention;

[0021] Figure 2 This is a schematic diagram of intermittent sampling repeated forwarding interference provided by an embodiment of the present invention;

[0022] Figure 3 It is a flow chart of a multi-channel main lobe interference separation and velocity information extraction method based on MIMO radar provided by an embodiment of the present invention;

[0023] Figure 4 is a schematic diagram of a MIMO radar array element diagram provided by an embodiment of the present invention;

[0024] Figure 5 is a time-frequency diagram of an echo signal provided by an embodiment of the present invention;

[0025] Figure 6 It is the result of the echo signal provided by the embodiment of the present invention after pulse compression processing;

[0026] Figure 7is a schematic diagram of a first demixed signal obtained by the JADE blind source separation algorithm provided by an embodiment of the present invention;

[0027] Figure 8 is a schematic diagram of a second demixed signal obtained by the JADE blind source separation algorithm provided by an embodiment of the present invention;

[0028] Fig. 9 is a schematic diagram of a third demixed signal obtained by the JADE blind source separation algorithm provided by an embodiment of the present invention;

[0029] Fig.10 It is the result of coherent accumulation of echo signals before interference separation provided by the embodiment of the present invention;

[0030] Fig.11 is a distance-velocity two-dimensional image after interference separation provided by an embodiment of the present invention;

[0031] Fig.12 It is a structural block diagram of a multi-channel main lobe interference separation and velocity information extraction device based on MIMO radar provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0032] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0033] A first aspect of the present invention provides a multi-channel main lobe interference separation and velocity information extraction method based on MIMO radar.

[0034] Before introducing the method of the present invention, the MIMO radar signal model and the interference signal model involved in the present invention are first introduced.

[0035] Generally speaking, the radar signal source transmits a linear frequency modulation signal s LFM (t), the mathematical model can be described by the following expression:

[0036]

[0037] Where A is the amplitude of the signal; f0 is the carrier frequency; T is the signal time width; let B be the signal bandwidth, then K = B / T is the linear modulation frequency, also known as the frequency modulation slope, which represents the rate of change of frequency over time; rect(t / T) is the rectangular function, defined as:

[0038]

[0039] Intermittent sampling and forwarding jamming is a technique that interferes with radar detection by sampling, delaying, and forwarding radar signals. According to how the jamming signal is processed, intermittent sampling and forwarding jamming can be divided into two main forms: intermittent sampling direct forwarding jamming and intermittent sampling repeated forwarding jamming.

[0040] Assuming that the sampling signal of direct forwarding interference is p(t), then

[0041]

[0042] Where, T w represents the length of the intermittent sampling window, Represents the convolution operation, Q = T p / T q Indicates the number of sampling times of intermittent sampling directly forwarding interference in the current pulse, T p Represents the duration of the radar signal in one pulse cycle, T q represents the time interval between each intermittent sampling, and τ represents the sampling delay of the jammer. For the jth sampling period, the directly forwarded interference signal can be expressed as:

[0043] J d (t) = A j s(tT w )p(t) (4);

[0044] In the formula, A j is the amplitude of the interfering signal.

[0045] Figure 1 The figure shows a schematic diagram of intermittent sampling direct forwarding interference. It can be seen that it samples the intercepted transmission signal and forwards it only once, while intermittent sampling repeated forwarding interference will forward the same sampled signal multiple times.

[0046] For intermittent sampling repeated forwarding interference, each sampling signal will be repeatedly forwarded M times, where The interference signal can be expressed as:

[0047]

[0048] Intermittent sampling repeated forwarding interference such as Figure 2 shown.

[0049] When the radar system is working and is interfered by intermittent sampling and forwarding, the received nth pulse echo signal x(t,n) can be expressed as

[0050] x(t,n)=s(t,n)+J(t)+n(t) (6);

[0051] Where x(t,n) represents the target echo signal corresponding to the nth pulse, which represents the echo reflected by the actual target detected by the radar; J(t) represents the intermittent sampling forwarding interference signal received by the radar; and n(t) is the additive white Gaussian noise, which represents the random noise component in the radar receiving process.

[0052] The following is a detailed description of the multi-channel main lobe interference separation and velocity information extraction method based on MIMO radar provided by the present invention. Figure 3 , Figure 3 1 is a flow chart of a method for multi-channel main lobe interference separation and velocity information extraction based on MIMO radar provided by an embodiment of the present invention. The method mainly includes the following steps:

[0053] Step 1: In each pulse repetition period, the echo signal of the MIMO radar multi-channel is obtained, and pulse compression processing is performed to obtain the echo signal after pulse compression.

[0054] Specifically, Figure 4 As shown in the figure, in a MIMO radar system, it is assumed that there are M receiving array elements forming a uniform linear array, and the first array element is taken as the reference array element. In each pulse repetition period, the radar transmit signal is:

[0055]

[0056] Where u(t)(t∈[0,T]) is the radar’s transmitted waveform within one pulse repetition period, a0 is the amplitude of the transmitted signal, and f0 is the radar’s initial carrier frequency.

[0057] Assuming that the spacing between the receiving array elements is half a wavelength, that is, d = λ2 = c / 2f0, and the reference distance between the target and the radar is R0, and the angle is θ, then within one pulse repetition period, the MIMO radar receiving signal is:

[0058]

[0059] After coherent detection, the intermediate frequency signal is:

[0060]

[0061] In the formula,

[0062] It can be seen from equation (9) that the difference between the echoes of different receiving array elements is caused by the term exp(-j2πf0·m·dsinθ), that is, the difference between two adjacent receiving array elements is exp(-j2πf0dsinθ).

[0063] Assume that the jammer and the real target have different arrival angles and there is aliasing in the time domain and frequency domain. If there are K signal sources in space (including target signal source and jammer), let the echo signal of each signal source in a receiving array element be s r,k (t), k = 1, 2, ..., K. Assume that the signals of M array elements are received, the received signal in the mth array element can be expressed as:

[0064]

[0065] Where n m (t) is the noise signal, r k (t) is the signal to be separated. The mixed signal received in the M array elements is written as a mixed equation:

[0066] X r (t) = AR (t) + N (t) (11);

[0067] In the formula,

[0068]

[0069] R(t)=[r1(t),r2(t),…,r K (t)] T (13);

[0070] N(t)=[n1(t),n2(t),…n M (t)] T (14);

[0071] Considering that the actual multi-channel received signals are seriously aliased in the time domain, pulse compression is performed on them before separation, which can improve the separation effect to a certain extent. Pulse compression is performed on the received signals of each channel, and the signal after pulse compression is Y(t). Therefore, the mixed equation of the MIMO radar multi-channel received signal after pulse compression can be expressed as:

[0072] Y(t)=AP(t)+N h (t) (15);

[0073] P(t)=[P1(t),P2(t),…,P K (t)] T (16);

[0074] Where, P1(t)=r1(t)*h(t),…,P K (t) = r K (t)*h(t).

[0075] N h (t) = [N h1 (t),N h2(t),…,N hN (t)] T (17);

[0076] Where N h1 (t)=n1(t)*h(t),…,N hM (t) = n M (t)*h(t). Where h(t) represents the conjugate transpose of the reference signal and * is the convolution symbol.

[0077] The steering matrix A of the MIMO multi-channel receiving pulse compression signal model has the characteristics of a Vandermonde matrix, so A is a column-full rank matrix with a rank of M. The number of signal sources can be estimated by eigendecomposition and effectively separated using methods such as blind source separation.

[0078] This embodiment performs pulse compression processing on the collected echo signal to reduce the sensitivity of the JADE algorithm to noise, thereby improving the robustness of signal separation.

[0079] Step 2: Use the MDL algorithm to process the echo signal after pulse compression and estimate the number of signal sources.

[0080] Specifically, assuming that there are M receiving array elements in the radar system, after transmitting a linear frequency modulation signal, the echo signal received by the mth receiving array element in the nth pulse repetition period can be expressed as:

[0081]

[0082] Among them, K is the number of signal sources (referred to as signal sources), α i is the attenuation coefficient of the ith source, s(t) is the transmitted linear frequency modulation signal; τ i is the delay of the signal going to and from the target, n m (t) is the noise received by the mth receiving element.

[0083] Let the matched filter of the transmitted signal be denoted as h(t), which is in general form the time-reversal and conjugate of the transmitted signal:

[0084] h(t)=s * (Tt) (19);

[0085] The pulse compression signal after matched filtering is expressed as:

[0086]

[0087] For a MIMO radar with M array elements and N* pulse repetition periods, the echo signals received by the M array elements in each pulse repetition period are organized into a matrix:

[0088]

[0089] Among them, Y n It is the echo signal of M array elements in the nth pulse period.

[0090] In each pulse repetition period, the received echo signal is pulse compressed and the number of signal sources is estimated using the MDL algorithm. Step 2 can be implemented in the following sub-steps:

[0091] 21) Calculate the covariance matrix of the echo signals received by the M array elements in each pulse repetition period.

[0092] Specifically, for each group of received signals Y n , its covariance matrix can be expressed as:

[0093]

[0094] in, is the outer product of the echo signal, and N represents the number of sampling points.

[0095] 22) Perform eigenvalue decomposition on the covariance matrix, and sort the multiple eigenvalues ​​obtained by the decomposition according to their sizes to obtain a sorted eigenvalue set.

[0096] Through R Y Perform eigenvalue decomposition and get:

[0097] R Y =UΛU H (twenty three);

[0098] Where Λ is the eigenvalue matrix of the covariance matrix, and U is the corresponding eigenvector matrix.

[0099] Sort the eigenvalues ​​from large to small to get the sorted eigenvalue set:

[0100] λ1≥λ2≥…≥λ M (twenty four);

[0101] Among them, the first K large eigenvalues ​​correspond to the source signal, and the following eigenvalues ​​are usually smaller and close to the noise level.

[0102] 23) Based on the MDL criterion, the number of signal sources is estimated using each eigenvalue in the eigenvalue set, and the number of signal sources that minimizes the MDL criterion value is selected as the estimated number of signal sources.

[0103] In this embodiment, the MDL algorithm estimates the number of information sources by minimizing the description length. The expression of the MDL criterion is:

[0104]

[0105] Where MDL(m) represents the MDL criterion value, m represents the number of signal sources to be estimated, and λ i represents the i-th eigenvalue in the eigenvalue set, M represents the number of receiving array elements, and N represents the number of sampling points.

[0106] The MDL criterion evaluates the criterion values ​​under different numbers of signal sources m and selects the m that minimizes the MDL criterion value as the estimated number of signal sources.

[0107] Step 3: Based on the estimated number of signal sources, the JADE blind source separation algorithm is used to separate the echo signal after pulse compression to obtain a separated signal.

[0108] 31) Perform whitening preprocessing on the echo signal after pulse compression to obtain a whitened signal.

[0109] Before blind source separation, the received signal is first whitened. The purpose of signal whitening is to convert the covariance matrix of the mixed signal into a unit matrix, thereby eliminating the correlation between signals and simplifying subsequent processing.

[0110] (a) Calculation of the covariance matrix

[0111] First, the covariance matrix of the received signal is calculated.

[0112] It can be understood that the covariance matrix calculation here is the same as step 21), that is, for the echo signal Y received by M array elements in each pulse repetition period n , its covariance matrix is:

[0113]

[0114] (b) Calculation of the whitening matrix

[0115] Next, the covariance matrix R Y Perform eigenvalue decomposition and construct a whitening matrix.

[0116] R X =UΛU H (27);

[0117] Among them, U is the eigenvector matrix of the covariance matrix, and Λ is the eigenvalue matrix of the covariance matrix.

[0118] Based on this decomposition, the whitening matrix W can be constructed:

[0119] W=Λ -1 / 2 U H (28);

[0120] The whitening matrix can process the original signal so that its covariance matrix becomes the unit matrix.

[0121] (c) Whitening treatment

[0122] The received signal Y n Perform whitening processing to obtain the whitened signal Z(t):

[0123] Z(t)=WY n (29);

[0124] At this point, the covariance matrix of the whitened signal becomes the identity matrix:

[0125] E{Z(t)Z H (t)}=I (30);

[0126] The whitened signal Z(t) retains the independence and non-Gaussian characteristics of the signal.

[0127] 32) Construct a fourth-order cumulant matrix for the whitened signal.

[0128] Specifically, the key to the JADE algorithm is to construct the fourth-order cumulant matrix of the signal. The fourth-order cumulant is a high-order statistic that has a zero response to Gaussian noise and can therefore be used to separate non-Gaussian components in the signal.

[0129] For the whitened signal Z(t), the fourth-order cumulant is defined as:

[0130]

[0131] In the formula, Z i ,Z j ,Z k ,Z l They respectively represent the components of the quaternion i, j, k, and l corresponding to the whitened signal, and E represents the expectation.

[0132] In simplified notation, multiple fourth-order cumulant matrices are usually used to characterize different signal components. For each quaternion (i, j, k, l), calculate the fourth-order cumulant matrix Q corresponding to the m'th signal source m' , 0<m'≤m, and blind source separation is achieved by jointly diagonalizing these matrices.

[0133] The JADE algorithm usually characterizes high-order statistical properties by constructing a matrix containing multiple fourth-order cumulants. The process of constructing the fourth-order cumulant matrix is ​​as follows: first, the fourth-order cumulants of the signal components of each receiving antenna are calculated; then these cumulants are organized into a matrix Q m' .

[0134] 33) The fourth-order cumulant matrix is ​​jointly diagonalized using an iterative method to find the optimal unmixing matrix that makes the fourth-order cumulant matrix as diagonal as possible.

[0135] In the JADE algorithm, the key step of signal separation is to calculate a set of fourth-order cumulant matrices Q m' Perform joint diagonalization. The purpose of joint diagonalization is to find an unmixing matrix B so that the fourth-order accumulation matrix Q m' Get as close to a diagonal matrix as possible in the unmixed coordinate system.

[0136] The goal of joint diagonalization is to find a matrix B such that:

[0137] BQ m' B H ≈diag(32);

[0138] That is, through the matrix B transformation, the fourth-order cumulant matrix becomes a diagonal matrix.

[0139] To achieve joint diagonalization, the JADE algorithm minimizes the following objective function:

[0140]

[0141] Among them, ||·|| F represents the Frobenius norm of the matrix. By minimizing this objective function, the unmixing matrix B that makes the fourth-order cumulant matrix as diagonal as possible is found, and m' represents different signal sources.

[0142] The objective function is optimized by iterative methods to find the optimal unmixing matrix B. Usually, gradient descent or Jacobi method is used for iterative solution of joint diagonalization.

[0143] 34) Project the whitened signal onto the optimal unmixing matrix to obtain the separated signal.

[0144] Specifically, after finding the optimal unmixing matrix B, the whitened signal Z(t) is projected onto the unmixing matrix to obtain the separated source signal, that is, the separated signal, which is expressed as:

[0145]

[0146] At this time, the separated source signal It is an independent signal extracted from a mixed signal, including the target signal and the interference signal.

[0147] Step 4: For the separated signals obtained from all pulse repetition periods, use the peak-to-sidelobe ratio technology to identify the target, and perform coherent accumulation on the identified target signals to extract the target's velocity information.

[0148] The intermittent sampling of radar signals can be regarded as the multiplication of the intermittent sampling pulse p(t) and the radar signal s(t). Assume that the sampling pulse width is T a , T s =(K+1)T a is the sampling pulse repetition interval, K∈{1,2,3,…} represents the number of repeated forwarding times, then the intermittent sampling interference signal can be expressed as:

[0149]

[0150] After matched filtering, the intermittent sampling forwarding interference can be expressed as:

[0151]

[0152] For intermittent sampling direct forwarding interference, after pulse compression, it can be expressed as:

[0153]

[0154] The pulse compression result of intermittent sampling repeated forwarding interference can be regarded as the sum of the intermittent sampling direct forwarding interference pulse compression result after time shift. Therefore, the intermittent sampling repeated forwarding interference after pulse compression can be expressed as:

[0155]

[0156] From equations (37) and (38), we can see that intermittent sampling and forwarding interference usually only samples and forwards part of the radar signal, and has a certain discontinuity in the time domain, which leads to an increase in the side lobe of the interference pulse compression result after the interference is pulse compressed. However, after the target signal is pulse compressed, the side lobe of its pulse compression result remains at a low level.

[0157] Based on this feature, the Peak Side Lobe Ratio (PSLR) can be used to distinguish targets from interference. The PSLR is the ratio of the maximum sidelobe intensity to the main lobe peak intensity, and the formula is:

[0158]

[0159] Where PSLR (dB) represents the peak sidelobe ratio, A side Indicates the maximum sidelobe intensity, A main Represents the main lobe peak intensity.

[0160] First, the peak-to-sidelobe ratio of the separated signal is measured, and the signal with a higher peak-to-sidelobe ratio is regarded as the interference signal, and the signal with the lowest peak-to-sidelobe ratio is regarded as the target signal.

[0161] Secondly, for the target signal, the coherent accumulation of the target signal is achieved by reconstructing the two-dimensional matrices of the fast time dimension and the slow time dimension, thereby extracting the target's speed information.

[0162] Specifically, after successfully identifying the target signal, the next step is to perform coherent accumulation processing on the target signals of all pulse repetition periods. This process involves reconstructing the two-dimensional matrix of the fast time dimension and the slow time dimension to extract the distance and speed information of the target. The fast time dimension reflects the time characteristics of each pulse and is directly related to the distance of the target; while the slow time dimension shows the changes in multiple pulse repetition periods, reflects the Doppler frequency of the target, and then provides the speed information of the target. In this way, the limitation that a single channel cannot extract target speed information in virtual multi-PRI (Pulse Repetition Interval) is effectively solved.

[0163] The multi-channel mainlobe interference separation and speed information extraction method based on MIMO radar provided by the present invention first obtains the echo signal of the multi-channel of the MIMO radar in each pulse repetition period and performs pulse compression processing; then uses the MDL algorithm to process the echo signal after pulse compression to estimate the number of signal sources; then uses the JADE blind source separation algorithm to separate the echo signal after pulse compression; finally uses the peak sidelobe ratio technology to identify the target signal of the separated signal, and repeats this operation in each pulse repetition period, and performs coherent accumulation on the identified target signal to extract the speed information of the target. The method uses the multi-channel characteristics of the MIMO radar system to effectively solve the problem of target speed information loss after separation.

[0164] The multi-channel main lobe interference separation and velocity information extraction method based on MIMO radar provided by the present invention is verified and explained through simulation experiments below.

[0165] The specific experimental parameters are set as follows: set the LFM signal bandwidth B = 40MHz, time width T = 8μs, sampling rate Fs = 160MHz, interference type is intermittent sampling direct forwarding interference and intermittent sampling repeated forwarding interference (sampling one to three), signal-to-noise ratio is set to 0dB, and interference-to-signal ratio is set to 40dB. The target position, speed, and angle are 4km, 35m / s, and 20° respectively; the intermittent sampling direct forwarding interference position, speed, and angle are 5km, 35m / s, and 23° respectively; the intermittent sampling repeated forwarding interference position, speed, and angle are 4km, 50m / s, and 25° respectively.

[0166] The time-frequency diagram of the collected original echo signal is as follows: Figure 5 As shown, it includes the target signal, the intermittent sampling direct forwarding interference signal and the intermittent sampling repeated forwarding interference signal. Figure 6The result of the echo signal after pulse compression processing is shown, where the black part represents the target signal, the red part is the intermittent sampling direct forwarding interference, and the blue part is the intermittent sampling repeated forwarding interference. It can be clearly seen from the figure that the target signal is completely submerged by the interference signal. Figures 7 to 9 The unmixed signal obtained by the JADE blind source separation algorithm is shown. After unmixing, a total of three signals are obtained. According to the measured PSLR value, Figure 7 The PSLR value of is the smallest, which is the target signal, and Figure 8 , 9 is an interference signal, and it can be seen from the waveform that Figure 8 Directly forward the interference signal for intermittent sampling, Fig. 9 Repeatedly forward the interference signal for intermittent sampling. Fig.10 The results of coherent accumulation of echo signals before interference separation are shown. It can be seen that the distance and speed information of the target cannot be extracted. After interference separation, the target pulse pressure echoes within multiple PRI cycles are accumulated to form a two-dimensional matrix of fast time dimension and slow time dimension, and the distance-speed two-dimensional information of the target is successfully extracted through coherent accumulation. The results are shown in Figure 1. Fig.11 As shown, it can be seen that it is consistent with the experimental setting parameters.

[0167] Therefore, the effectiveness of the present invention is verified.

[0168] Based on the same inventive concept, the second aspect of the present invention also provides a multi-channel main lobe interference separation and velocity information extraction device based on MIMO radar. Fig.12 , Fig.12 : is a structural block diagram of a multi-channel main lobe interference separation and velocity information extraction device based on MIMO radar provided by an embodiment of the present invention, the device comprising:

[0169] The pulse compression processing module is used to obtain the echo signal of the multi-channel of the MIMO radar in each pulse repetition period, and perform pulse compression processing to obtain the echo signal after pulse compression;

[0170] A signal source estimation module is used to process the echo signal after pulse compression using the MDL algorithm to estimate the number of signal sources;

[0171] A signal separation module is used to separate the echo signal after pulse compression by using the JADE blind source separation algorithm based on the estimated number of signal sources to obtain a separated signal;

[0172] The signal extraction module is used to identify the target using the peak-to-sidelobe ratio technology for the separated signals obtained from all pulse repetition periods, and to perform coherent accumulation on the identified target signals to extract the target's speed information.

[0173] It should be noted that the device of the embodiment of the present invention is a device for implementing the above-mentioned multi-channel main lobe interference separation and velocity information extraction method based on MIMO radar, and thus has the same or similar beneficial effects as the method embodiment. For the device embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0174] The above contents are further detailed descriptions of the present invention in combination with specific preferred embodiments, and it cannot be determined that the specific implementation of the present invention is limited to these descriptions. For ordinary technicians in the technical field to which the present invention belongs, several simple deductions or substitutions can be made without departing from the concept of the present invention, which should be regarded as falling within the protection scope of the present invention.

Claims

1. A multi-channel main lobe interference separation and velocity information extraction method based on MIMO radar, characterized in that: include: In each pulse repetition period, the multi-channel echo signal of the MIMO radar is obtained, and pulse compression processing is performed to obtain the echo signal after pulse compression; The echo signal after the pulse compression is processed by using the MDL algorithm to estimate the number of signal sources; Based on the estimated number of signal sources, the echo signal after the pulse compression is separated by using the JADE blind source separation algorithm to obtain a separated signal; For the separated signals obtained from all pulse repetition periods, the peak-to-sidelobe ratio technique is used for target identification, and the identified target signals are coherently integrated to extract the target's velocity information.

2. The multi-channel main lobe interference separation and velocity information extraction method based on MIMO radar according to claim 1 is characterized in that: The echo signal after the pulse compression is processed by using the MDL algorithm to estimate the number of signal sources, specifically including: Calculate the covariance matrix of the echo signals received by the M array elements in each pulse repetition period; Performing eigenvalue decomposition on the covariance matrix, and sorting multiple eigenvalues ​​obtained by the decomposition according to size to obtain a sorted eigenvalue set; Based on the MDL criterion, the number of signal sources is estimated using each eigenvalue in the eigenvalue set, and the number of signal sources that minimizes the MDL criterion value is selected as the estimated number of signal sources.

3. The multi-channel main lobe interference separation and velocity information extraction method based on MIMO radar according to claim 2 is characterized in that: The expression of the MDL criterion is: Where MDL(m) represents the MDL criterion value, m represents the number of signal sources to be estimated, and λ i represents the i-th eigenvalue in the eigenvalue set, M represents the number of receiving array elements, and N represents the number of sampling points.

4. The multi-channel main lobe interference separation and velocity information extraction method based on MIMO radar according to claim 1 is characterized in that: Based on the estimated number of signal sources, the echo signal after the pulse compression is separated using the JADE blind source separation algorithm to obtain a separated signal, including: Performing whitening preprocessing on the echo signal after the pulse compression to obtain a whitened signal; For the whitened signal, construct a fourth-order cumulant matrix; Using an iterative method to jointly diagonalize the fourth-order cumulant matrix, and find an optimal unmixing matrix that makes the fourth-order cumulant matrix as diagonal as possible; The whitened signal is projected onto the optimal unmixing matrix to obtain a separated signal.

5. The multi-channel main lobe interference separation and velocity information extraction method based on MIMO radar according to claim 4 is characterized in that: Performing whitening preprocessing on the echo signal after the pulse compression to obtain a whitened signal, including: Calculate the covariance matrix of the echo signals received by the M array elements in each pulse repetition period. The calculation formula is: In the formula, R Y represents the covariance matrix, Y n represents the echo signal received by M array elements in each pulse repetition period, represents the outer product of the signal, and N represents the number of sampling points; For the covariance matrix R Y Perform eigenvalue decomposition and construct a whitening matrix. The formula is expressed as: R Y =UΛU H ;W=Λ -1 / 2 U H ; Where W represents the whitening matrix, Λ represents the eigenvalue matrix of the covariance matrix, and U represents the eigenvector matrix of the covariance matrix; Based on the whitening matrix, the echo signals received by the M array elements in each pulse repetition period are processed to obtain a whitened signal, which is expressed as: Z(t)=WY n ; Where Z(t) represents the whitened signal.

6. The multi-channel main lobe interference separation and velocity information extraction method based on MIMO radar according to claim 4 is characterized in that: For the whitened signal, a fourth-order cumulant matrix is ​​constructed, including: For the whitened signal, the fourth-order cumulants are calculated, and all the fourth-order cumulants are combined into a matrix to obtain a fourth-order cumulant matrix; wherein the expression of the fourth-order cumulants is: In the formula, Cum(Z i ,Z j ,Z k ,Z l ) represents the fourth-order cumulant, Z i ,Z j ,Z k ,Z l They respectively represent the components of the quaternion i, j, k, and l corresponding to the whitened signal, and E represents the expectation.

7. The multi-channel main lobe interference separation and velocity information extraction method based on MIMO radar according to claim 4 is characterized in that: The optimization objective function for jointly diagonalizing the fourth-order cumulant matrix using an iterative method is: In the formula, represents the optimization objective function, B represents the unmixing matrix, ||·|| F represents the Frobenius norm of the matrix, Q m' represents the fourth-order cumulant matrix, and m' represents different signal sources.

8. The multi-channel main lobe interference separation and velocity information extraction method based on MIMO radar according to claim 1 is characterized in that: For the separated signals obtained from all pulse repetition periods, the peak-to-sidelobe ratio technology is used for target identification, and the identified target signals are coherently accumulated to extract the target's velocity information; including: Measuring the peak sidelobe ratio of the separated signal, taking the signal with a higher peak sidelobe ratio as the interference signal, and taking the signal with the lowest peak sidelobe ratio as the target signal; For the target signal, the two-dimensional matrix of the fast time dimension and the slow time dimension is reconstructed to achieve the coherent accumulation of the target signal, thereby extracting the speed information of the target.

9. The multi-channel main lobe interference separation and velocity information extraction method based on MIMO radar according to claim 8, characterized in that: The peak sidelobe ratio is the ratio of the maximum sidelobe intensity to the main lobe peak intensity, and the formula is expressed as: Where PSLR (dB) represents the peak sidelobe ratio, A side Indicates the maximum sidelobe intensity, A main Represents the main lobe peak intensity.

10. A multi-channel main lobe interference separation and velocity information extraction device based on MIMO radar, used to implement the multi-channel main lobe interference separation and velocity information extraction method based on MIMO radar according to any one of claims 1 to 9, characterized in that: include: The pulse compression processing module is used to obtain the echo signal of the multi-channel of the MIMO radar in each pulse repetition period, and perform pulse compression processing to obtain the echo signal after pulse compression; A signal source estimation module, used to process the echo signal after the pulse compression by using the MDL algorithm to estimate the number of signal sources; A signal separation module, used for separating the echo signal after the pulse compression by using a JADE blind source separation algorithm based on the estimated number of signal sources to obtain a separated signal; The signal extraction module is used to identify the target using the peak-to-sidelobe ratio technology for the separated signals obtained from all pulse repetition periods, and to perform coherent accumulation on the identified target signals to extract the target's speed information.

Images