Sub-block based space-time adaptive processing method for airborne broadband radar

Abstract

The application discloses a kind of based on sub-block division airborne broadband radar space-time adaptive processing method, comprising: based on different engineering application scene to radar echo data sub-block division, specifically divided into: radar echo data under the application scene of the more antenna element number is divided into sub-array;Radar echo data under the application scene of the more pulse number in a CPI is divided into sub-CPI;Radar echo data under the application scene that both the more antenna element number and the more pulse number in a CPI is divided into sub-array joint sub-CPI;After dividing, the echo data is synthesized to obtain the synthesized echo data;After the synthesized echo data is processed, envelope alignment is obtained after the distance time domain data;After the distance time domain data is processed, space-time adaptive processing is obtained after alignment output data.The method can reduce the amount of operation, to meet the demand of some engineering scene to real-time processing more.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of radars, and particularly relates to an airborne broadband radar space-time adaptive processing method based on sub-block division. BACKGROUND

[0002] In actual engineering, in order to obtain satisfactory target detection power and angle resolution capability, the number of array elements of a radar antenna is very large, and the number of accumulated pulses is also large. For a broadband radar, under such conditions, the calculation amount of envelope movement compensation and space-time processing is often difficult to bear. Therefore, in actual application, a dimension-reduced space-time adaptive processing (STAP) technology which can greatly reduce the calculation amount and the requirement for independent and identically distributed samples is usually used to process radar signals.

[0003] The traditional subarray division is essentially a space-domain dimension-reduced STAP technology, and many domestic and foreign scholars have carried out relevant research. For example, Delong et al. studied two kinds of non-uniform subarray structures, and proposed a principle that the center distance between adjacent subarrays constituting a subarray is coprime. Nickel proposed an irregular subarray division method based on taper function quantization, which can avoid the appearance of grating lobes and grating zeros, and form a directional diagram with low sidelobes. Wang et al. studied a broadband interference technology based on subarray-level STAP. Xu et al. proposed an equal-noise-power method to design a non-uniform adjacent subarray structure, which can achieve quasi-optimal adaptive processing performance and good adaptive directional pattern preservation effect, and the influence of grating lobe effect can be ignored. Xie et al. proposed a subarray division method based on a genetic algorithm, which can obtain a good antenna directional diagram without obvious grating lobes. Yu et al. studied the mathematical modeling problem of subarray optimization division, and designed a subarray optimization algorithm according to the mathematical model. Duan et al. proposed a 3D-STAP method for reducing degrees of freedom, which can significantly reduce the required training samples and calculation amount while maintaining the suboptimal clutter suppression performance.

[0004] However, the existing subarray division methods are mostly based on narrowband systems, and few of them are separately considered for broadband systems, the corresponding division principles are discussed, and the matching dimension-reduced STAP process is designed. In addition, the existing methods are mostly based on the number of array elements of a radar antenna for subarray division, but in actual application, not only the number of array elements of a radar antenna is large, but also the number of pulses in a CPI (Coherent Processing Interval) is often large, and the calculation of long CPI will also increase the system operation amount. SUMMARY

[0005] In order to solve the above problems existing in the prior art, the present application provides a sub-block division based airborne broadband radar space-time adaptive processing method.

[0006] In a first aspect, the present application provides a sub-block division based airborne broadband radar space-time adaptive processing method, comprising the following steps:

[0007] Step 1: sub-block division of radar echo data based on different engineering application scenarios, specifically including:

[0008] Sub-array division of radar echo data in an application scenario with a large number of antenna elements;

[0009] Sub-CPI division of radar echo data in an application scenario with a large number of pulses within a CPI;

[0010] Sub-array joint sub-CPI division of radar echo data in an application scenario with a large number of antenna elements and a large number of pulses within a CPI;

[0011] Step 2: sub-block synthesis of the divided echo data to obtain synthesized echo data;

[0012] Step 3: envelope alignment processing of the synthesized echo data to obtain aligned range-time domain data;

[0013] Step 4: space-time adaptive processing of the aligned range-time domain data to obtain output data.

[0014] In a second aspect, the present application provides a sub-block division based airborne broadband radar space-time adaptive processing system, comprising:

[0015] A sub-block division module for sub-block division of radar echo data based on different engineering application scenarios; the sub-block division module is specifically used for:

[0016] Sub-array division of radar echo data in an application scenario with a large number of antenna elements;

[0017] Sub-CPI division of radar echo data in an application scenario with a large number of pulses within a CPI;

[0018] Sub-array joint sub-CPI division of radar echo data in an application scenario with a large number of antenna elements and a large number of pulses within a CPI;

[0019] A synthesis module for sub-block synthesis of the divided echo data to obtain synthesized echo data;

[0020] an envelope alignment module, configured to perform envelope alignment processing on the synthesized echo data to obtain aligned range-time domain data;

[0021] a space-time adaptive processing module, configured to perform space-time adaptive processing on the aligned range-time domain data to obtain output data.

[0022] The present application has the following beneficial effects:

[0023] The method for space-time adaptive processing of airborne broadband radar based on sub-block division provided by the present application, on the one hand, expands the simple spatial sub-array division to the time domain sub-CPI division and the space-time two-dimensional sub-array joint sub-CPI division according to the principle of traditional sub-array division and in combination with the characteristics of the broadband system, so that the influence of envelope movement can be ignored within the divided sub-blocks, and the degree of freedom of the broadband system is reduced; on the other hand, the dimension reduction STAP process is redesigned for the broadband system, and envelope alignment processing is added, and the amount of computation of envelope alignment and STAP is proportional to the degree of freedom of the system, so that the amount of computation can be reduced by the method, thereby meeting the demand for real-time processing in some engineering scenarios.

[0024] The present application will be further described in detail below with reference to the accompanying drawings and embodiments. BRIEF DESCRIPTION OF DRAWINGS

[0025] Figure 1 A flowchart of a method for space-time adaptive processing of airborne broadband radar based on sub-block division provided by an embodiment of the present application;

[0026] Figure 2 A schematic diagram of uniform and non-uniform sub-block division provided by an embodiment of the present application;

[0027] Figure 3 A schematic diagram of sub-array joint sub-CPI division provided by an embodiment of the present application;

[0028] Figure 4 A block diagram of a system for space-time adaptive processing of airborne broadband radar based on sub-block division provided by an embodiment of the present application;

[0029] Figure 5 A simulation result diagram of the spatial adaptive pattern of sub-array non-division, uniform division and non-uniform division under scenario 1 by using the method of the present application;

[0030] Figure 6 A simulation result diagram of the improvement factor curve of sub-array non-division, uniform division and non-uniform division under scenario 1 by using the method of the present application;

[0031] Figure 7 A simulation result diagram of the time domain adaptive pattern of sub-CPI non-division, uniform division and non-uniform division under scenario 2 by using the method of the present application;

[0032] Figure 8 Fig. 4 is a simulation result diagram of an improvement factor curve under the condition of no sub-CPI division, uniform division and non-uniform division of sub-CPI by using the method of the present application in scenario 2;

[0033] Figure 9 Fig. 5 is a simulation result diagram of a spatial adaptive direction pattern under the condition of no division, uniform space-time division, uniform space division, non-uniform time division, non-uniform space division and uniform time division and non-uniform space-time division by using the method of the present application in scenario 3;

[0034] Figure 10 Fig. 6 is a simulation result diagram of a time adaptive direction pattern under the condition of no division, uniform space-time division, uniform space division, non-uniform time division, non-uniform space division and uniform time division and non-uniform space-time division by using the method of the present application in scenario 3;

[0035] Figure 11 Fig. 7 is a simulation result diagram of an improvement factor curve under the condition of no division, uniform space-time division, uniform space division, non-uniform time division, non-uniform space division and uniform time division and non-uniform space-time division by using the method of the present application in scenario 3. DETAILED DESCRIPTION

[0036] The present application will be further described in connection with specific embodiments, but the embodiments of the present application are not limited thereto.

[0037] Embodiment 1

[0038] Please refer to Figure 1 , Figure 1 Fig. 1 is a flow diagram of an airborne broadband radar space-time adaptive processing method based on sub-block division provided by an embodiment of the present application, which comprises the following steps:

[0039] Step 1: Sub-block division of radar echo data based on different engineering application scenarios.

[0040] In this embodiment, the following three different engineering application scenarios are set:

[0041] Scenario 1: The number of antenna elements is large, the maximum moving distance of target echo signal between the elements exceeds 1 / 10 of the distance resolution, the number of pulses within a coherent processing time is small, the maximum moving distance of target echo signal between the pulses does not exceed 1 / 10 of the distance resolution, and only sub-array division is needed at this time.

[0042] Scenario 2: The number of pulses within a CPI is large, the maximum moving distance of target echo signal between the pulses exceeds 1 / 10 of the distance resolution, the number of antenna elements is small, the maximum moving distance of target echo signal between the elements does not exceed 1 / 10 of the distance resolution, and only sub-CPI division is needed at this time.

[0043] Scenario 3: The number of antenna array elements and the number of pulses within a single CPI are both large, and the maximum travel distance of the target echo signal between array elements and between pulses exceeds 1 / 10 of the range resolution. In this case, it is necessary to perform subarray joint sub-CPI division.

[0044] Step 1 can be specifically divided into:

[0045] Subarray division of radar echo data in application scenarios with a large number of antenna array elements.

[0046] Sub-CPIs are used to divide radar echo data in application scenarios with a large number of pulses within a CPI.

[0047] For radar echo data in application scenarios where both the number of antenna array elements and the number of pulses within a CPI are relatively large, subarray joint subCPI division is performed.

[0048] Specifically, the division of radar echo data subarrays in scenario 1 can be carried out according to the following process.

[0049] First, establish the first division principle.

[0050] Optionally, in this embodiment, the first division principle mainly includes the following five points:

[0051] a. From the perspective of the array pattern, grating lobes should be avoided as much as possible, while the sidelobe level should be kept as low as possible;

[0052] b. From the perspective of space-time adaptive processing, it should be ensured that the system degrees of freedom after dimensionality reduction meet the requirements of clutter suppression, so that the signal-to-clutter ratio performance loss caused by dimensionality reduction is as low as possible;

[0053] c. It is necessary to minimize system costs and reduce the computational load of envelope alignment and space-time adaptive processing;

[0054] d. The beam pointing offset within each subarray must be controlled within the allowable range;

[0055] Because broadband systems have a wide frequency range, and the phase shift values ​​of each array element are frequency-dependent, the beam pointing will be offset due to the phase shift value calculated using the carrier frequency. To reduce the impact of beam pointing offset, the aperture size of each subarray must satisfy the following formula:

[0056]

[0057] Among them, D z θ represents the aperture size of the subarray, c represents the speed of light, B represents the signal bandwidth, and θ0 represents the maximum scanning angle of the beam direction.

[0058] e.The maximum aperture transit time in the subarray should be much smaller than the time resolution of the system.

[0059] In order to make the envelope movement between the elements in the divided subarray negligible, so as to avoid waveform distortion after beam forming, the maximum aperture transit time should be much smaller than the time resolution of the system, that is, the following formula should be satisfied:

[0060]

[0061] Then, the radar echo data is divided in a uniform division or non-uniform division manner based on the first division principle.

[0062] Please refer to Figure 2 , Figure 2 The schematic diagrams of the uniform division and non-uniform division of the subblock provided by the embodiment of the application are shown in the figures (a) and (b).

[0063] When the uniform division manner is adopted, assuming that the array is a uniform linear array composed of N elements, the array is uniformly divided into P subarrays, the number of elements in each subarray is N0, the signal wavelength is λ, and the distance between the elements is d, the formula for calculating the antenna pattern is:

[0064]

[0065] wherein, θ represents the included angle between the echo direction and the normal direction of the array, θ0 represents the pointing direction of the antenna pattern, and f2(θ) represents the array factor when the subarrays are combined.

[0066] Since the subarray division is performed, the distance between the array elements is changed from d to N0d, which has exceeded half a wavelength, and the distances are equal. If a certain direction angle θ1 satisfies the following equation, grating lobes will appear in the direction:

[0067]

[0068] It can be found from the above formula (4) that, as the number of elements N0 contained in the subarray increases, the angle interval of the grating lobes becomes shorter, and the number of the grating lobes increases.

[0069] When the non-uniform division manner is adopted, assuming that the array is a uniform linear array composed of N elements, the array is non-uniformly divided into P subarrays, the number of elements contained in the ith subarray is N i , and the formula for calculating the antenna pattern is:

[0070]

[0071] wherein, f2(i, θ) represents the array factor of the ith subarray, Δr irepresents the distance between the phase center of the ith subarray and the phase center of the reference subarray.

[0072] As can be seen from the above formula (5), the non-uniform division breaks the periodicity of the subarray arrangement, the number of elements of each subarray is different, and the corresponding array factor is also different, so as to possibly overcome the occurrence of the grating lobe phenomenon.

[0073] Further, for the radar echo data sub-CPI division under scenario 2, the following flow can be performed.

[0074] First, the second division principle is set.

[0075] Optionally, in the embodiment, the second division principle includes the following four points:

[0076] a. From the perspective of Doppler filtering, the grating lobe should be avoided as much as possible, and the sidelobe level should be as low as possible;

[0077] b. From the perspective of space-time adaptive processing, the system degrees of freedom after dimension reduction should be ensured to meet the requirements of clutter suppression, and the performance loss of signal-to-clutter ratio caused by dimension reduction should be as low as possible;

[0078] c. It is necessary to reduce the system cost as much as possible to reduce the computational complexity of envelope alignment and space-time adaptive processing;

[0079] d. It is necessary to meet the requirement that the maximum distance walk between pulses within a sub-CPI is much smaller than the two-way range resolution of the system.

[0080] In order to make the division of each sub-CPI, the maximum distance walk between pulses can be ignored, it is necessary to meet the requirement that the maximum distance walk between pulses is much smaller than the two-way range resolution of the system, that is, to meet the following formula:

[0081]

[0082] Wherein, CPI z represents the length of the sub-CPI, c represents the speed of light, B represents the signal bandwidth, v r represents the relative speed of the target and the carrier.

[0083] Then, the radar echo data is divided based on the second division principle in a uniform division or non-uniform division manner.

[0084] Specifically, the division manner of scenario 2 also includes uniform division and non-uniform division.

[0085] Please continue to see Figure 2When the uniform division method is adopted, assuming that the number of pulses transmitted in a CPI is M, the pulses are uniformly divided into Q shorter sub-CPIs, the number of pulses in each sub-CPI is M0, the signal wavelength is λ, and the pulse repetition period is T, the Doppler response function at this time is:

[0086]

[0087] where f d represents the Doppler frequency of a certain echo signal, f d0 represents the center frequency of the Doppler filter, and g2(f d ) represents a weighting factor when the sub-CPIs are combined.

[0088] After the sub-CPI division is performed, the slow time domain interval changes from T to M0T, and the intervals are equal. At this time, if the Doppler frequency f d1 of a certain signal satisfies the following equation, grating lobes will appear in the direction:

[0089]

[0090] It can be found by observing the above formula (8) that as the number of pulses M0 contained in the sub-CPI increases, the interval at which the grating lobes appear becomes shorter, and the number of grating lobes increases.

[0091] When the non-uniform division method is adopted, assuming that the number of pulses transmitted in a CPI is M, the pulses are non-uniformly divided into Q shorter sub-CPIs, the number of pulses contained in the i-th sub-CPI is M i , and the Doppler response function at this time is:

[0092]

[0093] where g2(i,f d ) represents the weighting factor of the i-th sub-CPI, represents the distance between the phase center of the i-th sub-CPI and the phase center of the reference sub-CPI.

[0094] It can be seen from the above formula (9) that the non-uniform division disrupts the periodicity of the arrangement of the sub-CPIs, the number of pulses in each sub-CPI is different, and the corresponding weighting factor is also different, so that it is possible to overcome the appearance of grating lobes.

[0095] Further, for the sub-array joint sub-CPI division of the radar echo data under scenario 3, the following process can be performed.

[0096] First, the third division principle is set.

[0097] In the embodiment, the third division principle only needs to combine the division principles of scene 1 and scene 2, that is, the third division principle includes the contents of the first division principle and the second division principle.

[0098] Then, the radar echo data is divided based on the third division principle in a manner of uniform space-time domain division, uniform space domain and non-uniform time domain division, non-uniform space domain and uniform time domain division or non-uniform space-time domain division.

[0099] Please refer to Figure 3 , Figure 3 The schematic diagram of the subarray joint sub-CPI division mode provided by the embodiment of the application is shown in FIG. 1. In the figure, (a) is uniform space-time domain division, (b) is uniform space domain and non-uniform time domain division, (c) is non-uniform space domain and uniform time domain division, and (d) is non-uniform space-time domain division. For the grating lobe effect of the space domain antenna pattern and the time domain Doppler filter caused by the uniform space-time domain division, reference can be made to the analysis of scene 1 and scene 2, and the embodiment will not be described in detail here.

[0100] Optionally, as an implementation manner, for the non-uniform division under scene 1, scene 2 and scene 3, an equal noise power method, a genetic algorithm or an ant colony algorithm can be used to realize, and the detailed process can be referred to the implementation of the related art.

[0101] In addition to dividing the array elements in the space domain, the application further extends to the time domain, divides the long CPI into sub-CPIs with fewer pulses, and further reduces the envelope alignment and the operation amount of STAP.

[0102] Step 2: Sub-block synthesis is performed on the divided echo data to obtain the synthesized echo data.

[0103] 21) Obtain the dimension reduction conversion matrix under different application scenes.

[0104] 22) Perform dimension reduction processing on the original echo data based on the dimension reduction conversion matrix to obtain the subarray synthesized echo data.

[0105] Specifically, for scene 1:

[0106] Suppose that the antenna array is a uniform linear array composed of N array elements, and is divided into P adjacent subarrays, and the number of array elements contained in each subarray is N1, N2, …, NP respectively, and satisfies P The subarray synthesis is actually equivalent to dimension reduction processing in the space domain, and the dimension reduction conversion matrix can be expressed as:

[0107] T s =T1T2T3(10)

[0108] ​Where T1 represents the amplitude weighting matrix at the element level, T2 represents the phase shifting matrix at the element level, and T3 shows the subarray partitioning method, which can be represented as follows:

[0109]

[0110]

[0111]

[0112] Where, ω i θ represents the amplitude weighting value of the i-th array element, θ0 represents the desired signal direction, and q represents the amplitude weighting value of the i-th array element. i (θ0) represents the phase shift value of the i-th array element.

[0113] Let X0 represent the original echo data, then the echo data synthesized by the subarray can be represented as:

[0114] X1 = T s X0 (14)

[0115] For scenario 2:

[0116] Suppose a CPI contains M pulses, and it is divided into Q adjacent sub-CPIs, each containing M1, M2, ..., M pulses respectively. Q ,satisfy The sub-CPI partitioning is essentially equivalent to dimensionality reduction in the time domain, and the dimensionality reduction transformation matrix can be represented as:

[0117] T t =T1T2T3 (15)

[0118] Where T1 represents the amplitude weighting matrix of each pulse, T2 represents the phase shift matrix of each pulse, and T3 shows the sub-CPI division method, which can be represented as follows:

[0119]

[0120]

[0121]

[0122] Where, ω i v represents the amplitude weighting value of the i-th pulse. r q represents the relative velocity between the target and the carrier aircraft. i (v r ) represents the phase shift value of the i-th pulse.

[0123] Let X0 represent the original echo data, then the echo data synthesized from the sub-CPI can be represented as:

[0124] X1 = T t X0 (19)

[0125] For scenario 3:

[0126] Suppose that the antenna array is a uniform linear array composed of N elements, the number of pulses in a CPI is M, and first, the pulses are divided into P adjacent sub-arrays, and the number of elements included in each sub-array is N1, N2, …, N P , satisfying Then, the sub-arrays are divided into Q adjacent sub-CPIs, and the number of pulses included in each sub-CPI is M1, M2, …, M Q , satisfying At this time, PQ sub-blocks are formed.

[0127] The dimension reduction conversion matrix of the sub-array combined with the sub-CPI division can be expressed as:

[0128] T = T s T t (20)

[0129] Wherein, T s represents the dimension reduction conversion matrix of the spatial domain, and T t represents the dimension reduction conversion matrix of the time domain.

[0130] Let X0 represent the original echo data, and the echo data after the sub-block synthesis can be expressed as:

[0131] X1 = TX0 (21)

[0132] Through the above process, the echo data after the sub-array synthesis X1 can be obtained.

[0133] Step 3: Perform envelope alignment processing on the synthesized echo data to obtain aligned range-time domain data.

[0134] Optionally, in the embodiment, the envelope alignment processing can be implemented by using any one of a keystone transformation interpolation method, a maximum correlation method or a minimum entropy method. The embodiment preferably uses the keystone transformation interpolation method to perform envelope alignment processing. Then, step 3 specifically includes:

[0135] 31) For different scenarios, the phase centers of each data sub-block are calculated.

[0136] Specifically, for scenario 1, first, the distance between the phase center of each sub-array and the phase center of the reference sub-array needs to be calculated from the element position and the weight. Since the amplitude weight values of the elements are different, the phase center of the i-th sub-array should be calculated by the following formula:

[0137]

[0138] Wherein, χij and ω ij respectively represent the coordinate and amplitude weighting value of the jth element of the ith subarray.

[0139] For scenario 2, first, the distance between the phase center of each sub-CPI and the phase center of the reference sub-CPI needs to be calculated by the slow-time coordinate and the weighting. The phase center of the ith sub-CPI is calculated by the following formula:

[0140]

[0141] where τ ij and ω ij respectively represent the slow-time coordinate and amplitude weighting value of the jth pulse of the ith sub-CPI.

[0142] For scenario 3, first, the distance between the phase center of each subarray and the phase center of the reference subarray needs to be calculated by the element position and the weighting, and the distance between the phase center of each sub-CPI and the phase center of the reference sub-CPI needs to be calculated by the slow-time coordinate and the weighting. The phase centers of each subarray and sub-CPI are calculated by formula (22) in the reference scenario 1 and formula (23) in scenario 2 respectively.

[0143] 32) Based on the phase center, the synthesized echo data is transformed into the range frequency domain by FFT transformation, and a spatial sinc interpolation processing is performed to obtain the spatially interpolated range frequency domain data.

[0144] Specifically, the synthesized echo data X1 is subjected to FFT transformation to obtain the sub-block synthesized range frequency domain echo data, denoted as X1(f l ,χ i ,τ).

[0145] Then, a spatial sinc interpolation processing is performed to obtain:

[0146]

[0147] where X2(f l ,χ',τ) represents the spatially interpolated range frequency domain data, f l represents the range frequency, f c represents the carrier frequency, χ i represents the array variable, which is the distance between the phase center of the ith subarray and the phase center of the reference subarray at this time, and τ represents the slow-time variable.

[0148] 33) The spatially interpolated range frequency domain data is subjected to a time-domain sinc interpolation processing to obtain the time-domain interpolated range frequency domain data, denoted as:

[0149]

[0150] where k = int(f d / f r ) represents the Doppler ambiguity number, f d represents the Doppler frequency of the target, f r represents the pulse repetition frequency, and T represents the pulse repetition period.

[0151] 34) Transform the distance-frequency domain data after time-domain interpolation to the distance time domain to obtain aligned distance time domain data, denoted as X3.

[0152] Step 4: Perform space-time adaptive processing on the aligned distance time domain data to obtain output data.

[0153] 41) Obtain the target space-time steering vector after sub-block synthesis under different scenarios.

[0154] Specifically, for scenario 1, the target space-time steering vector after sub-array synthesis is:

[0155]

[0156] For scenario 2, the target space-time steering vector after sub-CPI synthesis is:

[0157]

[0158] For scenario 3, the target space-time steering vector after sub-block synthesis is:

[0159]

[0160] where S s and S t represent the spatial steering vector and the time steering vector of the target signal, respectively.

[0161] 42) Estimate the covariance matrix using the maximum likelihood method, denoted as:

[0162]

[0163] where represents the distance time domain data of the i-th distance cell after alignment, and L represents the number of distance cells.

[0164] 43) Based on the target space-time steering vector and the covariance matrix, an optimization equation is established based on the principle of maximum output signal-to-clutter-and-noise ratio.

[0165] where the optimization equation is denoted as:

[0166]

[0167] 44) solving the optimization equation to obtain the optimal weight vector, and obtaining the output data of the airborne broadband radar space-time adaptive processing according to the optimal weight vector.

[0168] Specifically, by solving the optimization equation, the optimal weight vector is obtained as follows:

[0169]

[0170] Wherein, is a normalized constant.

[0171] Finally, the output of the STAP method of the airborne broadband radar based on sub-block division is obtained as follows:

[0172]

[0173] At this point, the space-time adaptive processing of the airborne broadband radar based on sub-block division is completed.

[0174] The space-time adaptive processing method of the airborne broadband radar based on sub-block division provided by the application, on the one hand, according to the principle of traditional sub-array division, combined with the characteristics of the broadband system, expands the pure spatial sub-array division to the time domain sub-CPI division, and the space-time two-dimensional sub-array joint sub-CPI division, so that the influence of envelope movement can be ignored in the sub-block after division, and the degree of freedom of the broadband system is reduced; on the other hand, the dimension reduction STAP process is redesigned for the broadband system, and the envelope alignment processing is increased. Since the operation amount of STAP processing is mainly concentrated in the estimation and inversion of the covariance matrix, and for the broadband system, the operation amount of envelope alignment also needs to be considered, and the operation amount of these two parts is proportional to the degree of freedom of the system, therefore, the method can save the operation amount of envelope alignment and STAP, so as to meet the demand of real-time processing in some engineering scenes.

[0175] Embodiment two

[0176] On the basis of the above-mentioned embodiment one, the embodiment provides a space-time adaptive processing system of airborne broadband radar based on sub-block division. Please refer to Figure 4 , Figure 4 The structure block diagram of the space-time adaptive processing system of airborne broadband radar based on sub-block division provided by the embodiment of the application, the system comprises:

[0177] The sub-block division module is used for sub-block division of radar echo data based on different engineering application scenarios; the sub-block division module is specifically used for:

[0178] Sub-array division of radar echo data in the application scenario with more antenna elements;

[0179] The radar echo data in an application scene with a large number of pulses in a CPI is divided into sub-CPIs;

[0180] The radar echo data in an application scene with a large number of antenna elements and a large number of pulses in a CPI is divided into sub-arrays and sub-CPIs;

[0181] The synthesis module is configured to synthesize the divided echo data to obtain synthesized echo data;

[0182] The envelope alignment module is configured to perform envelope alignment processing on the synthesized echo data to obtain aligned range-time domain data;

[0183] The space-time adaptive processing module is configured to perform space-time adaptive processing on the aligned range-time domain data to obtain output data.

[0184] The system provided in this embodiment can be used to implement the method provided in the first embodiment, and the detailed process can be referred to the description of the first embodiment. Therefore, the system can also reduce the amount of calculation, thereby better meeting the demand for real-time processing in some engineering scenarios.

[0185] Embodiment three

[0186] The effectiveness of the present application is verified and described below through simulation experiments.

[0187] Experiment 1: Simulation of scenario 1

[0188] The simulation conditions are as follows: the antenna array is a uniform linear array containing 60 elements, the element spacing is 0.15 m, the number of pulses in one coherent processing interval is 5, the pulse repetition frequency is 8000 Hz, the signal carrier frequency is 1 GHz, and the bandwidth is 100 MHz. The three ways of no division, uniform division, and non-uniform division are all added with a 35dB Chebyshev window at the element level; among them, the uniform division is divided into 30 sub-arrays, each containing 2 elements, and the non-uniform division is also divided into 30 sub-arrays, and the element numbers of each sub-array are obtained by using the equal noise method, which are 10, 4, 2, 2, 1, 2, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 2, 2, 2, 4, 10; whether uniform division or non-uniform division, each sub-array meets the conditions of controllable beam pointing offset and negligible envelope walking; the three ways have all been subjected to envelope alignment processing before STAP processing.

[0189] The simulation results are shown in Figure 5 and Figure 6 .

[0190] Figure 5The simulation result diagram of the space domain adaptive pattern of the subarray under the scene 1 and the non-partition, the uniform partition and the non-uniform partition by using the method of the present application is shown in the figure. It can be seen that the grating lobe appears obviously after the uniform partition of the subarray, and the sidelobe level is high, which can easily cause false alarm and is not conducive to target detection; only a shallow notch is formed at the clutter position, and the grating zero point is formed at many positions, and the clutter suppression capability is poor. The grating lobe and the grating zero point do not appear in the non-uniform partition, compared with the non-partition, the sidelobe level is slightly higher, the clutter notch is slightly shallower, and the STAP performance is lost to some extent, but the adaptive pattern is good in general, and basically has the ability of clutter suppression and target detection.

[0191] Figure 6 The simulation result diagram of the improvement factor curve of the subarray under the scene 1 and the non-partition, the uniform partition and the non-uniform partition by using the method of the present application is shown in the figure. It can be seen that the improvement factor of the uniform partition and the non-uniform partition of the subarray is decreased compared with the non-partition; this is because the degree of freedom of the system is decreased after the partition of the subarray. In the sidelobe area, the improvement factor of the non-uniform partition is obviously better than that of the uniform partition. In the main lobe area, the notch of the non-uniform partition is slightly narrower, and the target detection performance is slightly better, but it is slightly wider than the non-partition; this is because the uniform partition produces the grating lobe and the grating zero point, and part of the grating zero point may enter the main lobe, which causes the distortion of the space domain adaptive pattern, and the STAP performance is decreased. Compared with the full-dimensional adaptive processing, the non-uniform partition only sacrifices a small part of the clutter suppression and MDV performance, but saves a large amount of envelope alignment and STAP operation amount due to the decrease of the system degree of freedom.

[0192] Experiment 2: Simulation of scene 2

[0193] The simulation conditions are as follows: the antenna array is a uniform linear array containing 5 array elements, the array element spacing is 0.15 m, the number of pulses in a coherent processing interval is 60, the pulse repetition frequency is 8000 Hz, the signal carrier frequency is 1 GHz, and the bandwidth is 100 MHz. The three ways of non-partition, uniform partition and non-uniform partition of the sub-CPI are all in the Chebyshev window of 75 dB at each pulse; among them, the uniform partition is divided into 20 sub-CPIs, each sub-CPI contains 3 pulses, the non-uniform partition is also divided into 20 sub-CPIs, and the number of pulses of each sub-CPI is obtained by using the equal noise method, which is 18, 3, 2, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 2, 2, 2, 18; whether the uniform partition or the non-uniform partition, each sub-CPI meets the condition of neglecting envelope movement; the three ways have been envelope aligned before STAP processing.

[0194] The simulation results are shown in Figure 7 and Figure 8 .

[0195] Figure 7 Figure 2 is a simulation result diagram of time-domain adaptive pattern under sub-CPI non-division, uniform division and non-uniform division in scene 2 using the method of the present application. It can be seen that after uniform division of the sub-CPI, two obvious grating lobes appear, which can easily cause false alarm and is not conducive to target detection; only a shallow notch is formed at the clutter position, and grating nulls are formed at many positions, and the clutter suppression capability is poor. The non-uniform division of the sub-CPI does not appear grating lobes and grating nulls, and compared with the non-division, the sidelobe level is slightly higher, the clutter notch is slightly shallower, and the STAP performance has some loss, but overall the adaptive pattern is well preserved, and basically has the ability of clutter suppression and target detection.

[0196] Figure 8 Figure 2 is a simulation result diagram of time-domain adaptive pattern under sub-CPI non-division, uniform division and non-uniform division in scene 2 using the method of the present application. It can be seen that after uniform division of the sub-CPI, two obvious grating lobes appear, which can easily cause false alarm and is not conducive to target detection; only a shallow notch is formed at the clutter position, and grating nulls are formed at many positions, and the clutter suppression capability is poor. The non-uniform division of the sub-CPI does not appear grating lobes and grating nulls, and compared with the non-division, the sidelobe level is slightly higher, the clutter notch is slightly shallower, and the STAP performance has some loss, but overall the adaptive pattern is well preserved, and basically has the ability of clutter suppression and target detection.

[0197] Experiment 3: Simulation of scene 3

[0198] The simulation conditions are as follows: the antenna array is a uniform linear array containing 36 array elements, the array element spacing is 0.15 m, the number of pulses in a coherent processing interval is 36, the pulse repetition frequency is 8000 Hz, the signal carrier frequency is 1 GHz, and the bandwidth is 100 MHz. The five ways of non-division, space-time domain uniform division, space domain uniform time domain non-uniform division, space domain non-uniform time domain uniform division and space-time domain non-uniform division are all added with a 30 dB Chebyshev window at each array element, and each pulse is added with a 60 dB Chebyshev window; wherein the subarray uniform division is divided into 18 subarrays, each subarray contains 2 array elements, the non-uniform division is also divided into 18 subarrays, and the array element number of each subarray is obtained by using the equal noise method, which is 7, 2, 2, 2, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 3, 3, 6; the sub-CPI uniform division is divided into 12 sub-CPIs, each sub-CPI contains 3 pulses, the non-uniform division is also divided into 12 sub-CPIs, and the pulse number of each sub-CPI is obtained by using the equal noise method, which is 11, 2, 2, 1, 1, 1, 1, 1, 1, 2, 2, 11; the four division modes finally form 216 subblocks, each subblock meets the condition that the envelope movement between array elements and pulses is ignored, and the beam pointing offset is controlled within the allowable range; the envelope alignment processing is performed before the STAP processing of the five ways.

[0199] The simulation results are shown in Figures 9-11 .

[0200] Figure 9 The simulation results of the space domain adaptive pattern in the scene 3 under the conditions of non-division, space-time domain uniform division, space domain uniform time domain non-uniform division, space domain non-uniform time domain uniform and space-time domain non-uniform division by using the method of the present application are shown in Figure 10The time domain self-adapting pattern simulation result diagram of the method of the present application under scene 3 without partition, space-time domain uniform partition, space domain uniform time domain non-uniform partition, space domain non-uniform time domain uniform and space-time domain non-uniform partition. It can be seen that the space-time domain uniform partition appears obvious grating lobe in space and time domain pattern, the sidelobe level is higher, which easily causes false alarm; only a shallow notch is formed in the clutter position, and grating zero point is formed in many positions, the clutter suppression ability is poor. The space domain uniform time domain non-uniform partition only appears grating lobe in space domain self-adapting pattern, the time domain self-adapting pattern does not appear grating lobe, and the time domain sidelobe level is lower than that of space-time domain uniform partition. The space domain non-uniform time domain uniform partition only appears grating lobe in time domain self-adapting pattern, the space domain self-adapting pattern does not appear grating lobe, and the space domain sidelobe level is lower than that of space-time domain uniform partition. The space-time domain non-uniform partition does not appear grating lobe in space and time domain self-adapting pattern, and the overall sidelobe level is also lower than the above three cases, the STAP performance is the best; compared with the case without partitioning sub-block, the sidelobe level is slightly higher, the clutter notch is slightly shallow, and the STAP performance has certain loss, but overall the space and time domain self-adapting pattern is good, and basically has the ability of clutter suppression and target detection.

[0201] Figure 11 The improvement factor curve simulation result diagram of the method of the present application under scene 3 without partition, space-time domain uniform partition, space domain uniform time domain non-uniform partition, space domain non-uniform time domain uniform and space-time domain non-uniform partition. It can be seen that compared with the case without partition, the improvement factors of the four partition modes are decreased, which is because the freedom of the system is decreased after the sub-block partition. In the sidelobe area, the space-time domain uniform partition and the space domain non-uniform time domain uniform partition appear grating notch, which is because the Doppler response of the two partition modes appears grating lobe, which causes the STAP self-adapting to eliminate part of the target signal; the improvement factor of the space-time domain non-uniform partition is obviously better than the other three partition modes, and the improvement factor of the space-time domain uniform partition is the worst. In the main lobe area, the notch of the space-time domain non-uniform partition is narrower than the other three partition modes, and the target detection performance is better, but it is slightly wider than the case without partition; this is because the other three partition modes involve uniform partition, which produces grating lobe and grating zero point, and part of the grating zero point may enter the main lobe, causing the distortion of the space or (and) time domain self-adapting pattern, and the STAP performance is decreased. Compared with full-dimensional adaptive processing, the space-time domain non-uniform partition only sacrifices a small part of the clutter suppression and MDV performance, but saves a large amount of envelope alignment and STAP operation amount due to the decrease of the system freedom.

[0202] In summary, the effectiveness of the method of the present application in saving operation amount is verified.

[0203] The above is further detailed description of the present application in combination with specific preferred embodiments, and cannot be deemed as limitation of the specific implementation of the present application to these descriptions. For those skilled in the art to which the present application belongs, without departing from the concept of the present application, a number of simple deductions or substitutions can be made, and all should be deemed as falling within the protection scope of the present application.

Claims

1. A method for space-time adaptive processing of airborne broadband radar based on sub-block partitioning, characterized in that, The method comprises the following steps: Step 1: sub-block division is performed on radar echo data based on different engineering application scenarios, and the division is specifically as follows: For radar echo data in an application scenario with a large number of antenna elements, sub-array division is performed; For radar echo data in an application scenario with a large number of pulses in one CPI, sub-CPI division is performed; For radar echo data in an application scenario with a large number of both antenna elements and pulses in one CPI, sub-array joint sub-CPI division is performed; Step 2: sub-block synthesis is performed on the divided echo data to obtain synthesized echo data; Step 3: envelope alignment processing is performed on the synthesized echo data to obtain aligned range-time domain data; Step 4: space-time adaptive processing is performed on the aligned range-time domain data to obtain output data.

2. The sub-block division based airborne wideband radar space-time adaptive processing method according to claim 1, characterized in that, In step 1, for radar echo data in an application scenario with a large number of antenna elements, sub-array division is performed, including: A first division principle is set for an application scenario with a large number of antenna elements in a wideband system; Based on the first division principle, the radar echo data is divided in a uniform division or non-uniform division manner; wherein the first division principle comprises: a. From the perspective of the array pattern, grating lobes should be avoided as much as possible, and the sidelobe level should be as low as possible; b. From the perspective of space-time adaptive processing, the system degrees of freedom after dimension reduction should meet the requirements of clutter suppression, and the signal-to-clutter ratio performance loss caused by dimension reduction should be as low as possible; c. The system cost needs to be reduced as much as possible to reduce the computational complexity of envelope alignment and space-time adaptive processing; d. The beam pointing offset in each sub-array needs to be controlled within the allowable range; e. The maximum aperture transit time in the sub-array needs to be much smaller than the time resolution of the system.

3. The sub-block division based airborne wideband radar space-time adaptive processing method according to claim 1, characterized in that, In step 1, for radar echo data in an application scenario with a large number of pulses in one CPI, sub-CPI division is performed, including: A second division principle is set for an application scenario with a large number of pulses in one CPI in a wideband system; Based on the second division principle, the radar echo data is divided in a uniform division or non-uniform division manner; wherein the second division principle comprises: a. From the perspective of Doppler filtering, grating lobes should be avoided as much as possible, and the sidelobe level should be as low as possible; b. From the perspective of space-time adaptive processing, the system degrees of freedom after dimension reduction should meet the requirements of clutter suppression, and the signal-to-clutter ratio performance loss caused by dimension reduction should be as low as possible; c. The system cost needs to be reduced as much as possible to reduce the computational complexity of envelope alignment and space-time adaptive processing; d. The maximum distance migration between pulses in the sub-CPI needs to be much smaller than the two-way range resolution of the system.

4. The sub-block division based airborne wideband radar space-time adaptive processing method according to claim 1, characterized in that, In step 1, for radar echo data in an application scenario with a large number of both antenna elements and pulses in one CPI, sub-array joint sub-CPI division is performed, including: A third division principle is set for an application scenario with a large number of both antenna elements and pulses in one CPI in a wideband system; wherein the third division principle comprises the contents of the first division principle and the second division principle. The radar echo data is divided based on the third division principle in a manner of uniform space-time domain division, uniform space domain and non-uniform time domain division, non-uniform space domain and uniform time domain division, or non-uniform space-time domain division.

5. The subblock-based space-time adaptive processing method for airborne wideband radar according to any one of claims 2-4, characterized in that, The non-uniform division manner is implemented by using an equal noise power method, a genetic algorithm or an ant colony algorithm.

6. The sub-block division based airborne wideband radar space-time adaptive processing method according to claim 1, characterized in that, Step 2 comprises: 21) obtaining a dimension reduction conversion matrix under different application scenarios; 22) performing dimension reduction processing on the original echo data based on the dimension reduction conversion matrix to obtain echo data after subarray synthesis.

7. The sub-block division based airborne wideband radar space-time adaptive processing method according to claim 1, characterized in that, In step 3, the envelope alignment processing is implemented by using a keystone transform interpolation method, a maximum correlation method or a minimum entropy method.

8. The sub-block division based airborne wideband radar space-time adaptive processing method according to claim 1, characterized in that, Step 3 comprises: 31) calculating the phase center of each data subblock for different scenarios; 32) based on the phase center, the synthesized echo data is transformed to the range frequency domain by using FFT transform, and space domain sinc interpolation processing is performed to obtain space domain interpolated range frequency domain data; 33) performing time domain sinc interpolation processing on the space domain interpolated range frequency domain data to obtain time domain interpolated range frequency domain data; 34) transforming the time domain interpolated range frequency domain data to the range time domain to obtain aligned range time domain data.

9. The sub-block division based airborne wideband radar space-time adaptive processing method according to claim 1, characterized in that, Step 4 comprises: 41) obtaining a target space-time steering vector after subblock synthesis under different scenarios; 42) estimating a covariance matrix by using a maximum likelihood method; 43) based on the target space-time steering vector and the covariance matrix, an optimization equation is established based on the principle of maximum output signal-to-clutter-and-noise ratio; 44) solving the optimization equation to obtain an optimal weight vector, and obtaining output data of space-time adaptive processing of the airborne broadband radar according to the optimal weight vector.

10. A subblock-based space-time adaptive processing system for airborne broadband radar, comprising: Comprise: A subblock division module is configured to divide radar echo data based on different engineering application scenarios; the subblock division module is specifically configured to: perform subarray division on radar echo data under an application scenario with a large number of antenna elements; perform sub-CPI division on radar echo data under an application scenario with a large number of pulses in one CPI; perform subarray and sub-CPI joint division on radar echo data under an application scenario with a large number of antenna elements and a large number of pulses in one CPI; A synthesis module is configured to synthesize the divided echo data to obtain synthesized echo data; An envelope alignment module is configured to perform envelope alignment processing on the synthesized echo data to obtain aligned range time domain data; A space-time adaptive processing module is configured to perform space-time adaptive processing on the aligned range time domain data to obtain output data.

Images