Cognitive space-time adaptive radar

Abstract

The invention relates to the technical field of signal processing, and provides a cognitive space-time adaptive radar. The system comprises an identification module and a decision processing module. The identification module is used for identifying the type of the interference signal by using a convolutional neural network and identifying the type of the clutter signal according to the characteristic of each clutter signal; and the decision processing module is used for selecting a corresponding interference suppression method according to the type of the interference signal, selecting a corresponding clutter suppression method according to the type of the clutter signal, and processing the echo signal by using the interference suppression method and the clutter suppression method. According to the method, the type of the interference signal and the type of the clutter signal are identified, and the corresponding suitable processing method is selected according to the identified types, so that various types of interference and clutter can be dealt with, and the detection accuracy of the target in the complex environment is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of signal processing, and particularly to a cognitive space-time adaptive radar. Background Art

[0002] Relying on the altitude advantage of the platform, airborne radars not only overcome the shielding effects of the earth's curvature and terrain features, greatly expanding the detection range, but also can be flexibly deployed in actual use, providing valuable warning time and planning horizons for various operations. Due to the down-looking operation, airborne radars need to detect moving targets in the presence of strong ground / sea clutter. In particular, the high-speed movement of the platform causes the coupling of the clutter Doppler frequency and the echo direction, resulting in a severe broadening of its Doppler frequency, causing weak moving targets to be submerged in strong clutter signals in the frequency domain. Fortunately, with the increasing maturity of phased array antennas and pulse Doppler technologies, as well as the rapid improvement of hardware processing capabilities, airborne radars can perform two-dimensional joint signal processing in the spatial and temporal domains through the received multi-channel pulse echo data, significantly improving the detection performance of moving targets. Currently, space-time adaptive processing (STAP) technology has become one of the core key technologies of airborne radar systems.

[0003] With the continuous expansion of radar application requirements, airborne radars are facing increasingly complex terrain and electromagnetic environments. The weights of STAP are related to the echo data, so it can adapt to environmental changes. However, when the environment faced by the radar changes too quickly or the array form is a non-side-looking array, conformal array, end-fire array, or bistatic configuration, the radar echoes show a severely non-uniform and non-stationary distribution. Since there are not enough training samples that meet the independent and identically distributed conditions, the performance of STAP methods drops significantly. How to effectively detect weak moving targets in complex environments such as non-uniform clutter, non-stationary clutter, wind farm clutter, and dense unintentional interference is a severe challenge faced by airborne radars.

[0004] In view of this, overcoming the defects of the existing technology is an urgent problem to be solved in this technical field. Summary of the Invention

[0005] The technical problem to be solved by the present invention is to provide a cognitive space-time adaptive radar.

[0006] The present invention adopts the following technical solutions: In a first aspect, the present invention provides a cognitive space-time adaptive radar, including an identification module and a decision processing module; The identification module is used to identify the types of interference signals using a convolutional neural network and identify the types of clutter signals according to the characteristics of each clutter signal; The decision-making processing module is used to select a corresponding interference suppression method according to the type of interference signal, select a corresponding clutter suppression method according to the type of clutter signal, and process the echo signal using the interference suppression method and the clutter suppression method.

[0007] Preferably, the type of the clutter signal includes one or more of uniform clutter, non-uniform clutter, non-stationary clutter, and wind farm clutter; The selecting a corresponding clutter suppression method according to the type of the clutter signal specifically includes: When the type of the clutter signal is uniform clutter, select to use the STAP algorithm or the space-time polarization adaptive processing algorithm for clutter suppression; When the type of the clutter signal is non-uniform clutter, select to use the sample selection method based on geographic information to select uniform training samples, perform power difference correction on the selected uniform training samples, and based on the corrected uniform training samples, select the STAP algorithm or the color loading STAP method for clutter suppression; When the type of the clutter signal is non-stationary clutter, select to use the 3D-STAP method or the non-stationary clutter suppression method based on space-time clutter spectrum adaptive compensation for clutter suppression; When the type of the clutter signal is wind farm clutter, select to use the wind farm isolated point clutter suppression method based on micro-Doppler characteristics for suppression.

[0008] Preferably, the type of the interference signal includes one or more of unintentional interference, main lobe interference, side lobe interference, and main lobe unintentional interference; The selecting a corresponding interference suppression method according to the type of the interference signal specifically includes: When the type of the interference signal is unintentional interference or main lobe unintentional interference, select to use the airborne radar unintentional interference suppression method based on frequency domain adaptive filtering for interference suppression; When the type of the interference signal is main lobe interference, select to use the space-time polarization joint adaptive anti-main lobe interference method for interference suppression; When the type of the interference signal is side lobe interference, select to use the STAP method under the interference environment for interference suppression.

[0009] Preferably, the airborne radar unintentional interference suppression method based on frequency domain adaptive filtering specifically includes: Performing interference spectrum estimation through adaptive beamforming and frequency spectrum estimation algorithm based on the maximum likelihood criterion to obtain the power spectral density estimator of each frequency point; Determining the frequency range of the interference signal according to the power spectral density estimator of each frequency point; Construct multiple channels according to the frequency range of the interference signal, with each channel corresponding to an interference filter; In each channel, use the interference filter to filter out the interference signal to obtain the interference-free signal of each channel. For the color noise caused by filtering, perform noise compensation on the interference-free signal of each channel to obtain the channel output signal of each channel; Select the channel output signal with the minimum power as the final output signal.

[0010] Preferably, the interference spectrum is estimated by the adaptive beamforming and frequency spectrum estimation algorithm based on the maximum likelihood criterion to obtain the power spectral density estimator of each frequency point, specifically including: Determine that in the Gaussian white noise environment, the output signal corresponding to the u th pulse signal after adaptive beamforming at the corresponding spatial angle is ; where is the output signal of the interference signal of the u th pulse after adaptive beamforming; is the noise signal, and obeys the Gaussian distribution with a mean of 0 and a variance of ; After performing the mixed-radix FFT transform on , the observation model at the frequency point is ; where is the observed value of the noise signal intensity at the frequency u in the th observation, which obeys the complex Gaussian distribution with a mean of 0 and a variance of , that is ; is the observed value of the interference signal intensity at the frequency u in the th observation; K is the number of groups of received signals; When the array receives K groups of signals, the power spectral density estimator at the frequency point is calculated as ; where is the amplitude value estimator at the frequency point , ; obeys the complex Gaussian distribution with a mean of and a variance of , that is ; is the estimated value of the interference signal intensity at the frequency ; is the interference signal intensity at the frequency The estimated value of the noise signal intensity at the location.

[0011] Preferably, determining the frequency range of the interference signal according to the estimated power spectral density at each frequency point specifically includes: Based on the estimated power spectral density of the signal , establish a signal observation model as ; where represents the frequency point without interference, represents the frequency point with interference; According to the signal observation model, determine the estimated value of the interference-free power spectral density at the frequency point ; ; obeys an exponential distribution; where the estimated value of the interference-free power spectral density is the estimated value of the power spectral density corresponding to the case where there is no interference at the corresponding frequency point; Set the estimated value of the power spectral density at the frequency point as the detection unit, the estimated value of the power spectral density at the frequency point as the protection unit, the estimated value of the power spectral density at the frequency point as the reference unit, and determine the reference mean according to the estimated value of the power spectral density of the reference unit, and the expression is ; where is the estimated value of the power spectral density at the frequency point , is the estimated value of the power spectral density at the frequency point , is the length of the protection unit, is the length of the reference unit; According to the estimated value of the power spectral density at each frequency point and the reference mean , determine whether there is interference at each frequency point, and determine the frequency range of the interference signal according to the range of the frequency points with interference ; where, when the estimated value of the power spectral density at the frequency point is , it is determined that there is interference at the frequency point ; when the estimated value of the power spectral density at the frequency point is , it is determined that there is no interference at the frequency point ; where, P F is the false alarm probability determined according to historical experience, .

[0012] Preferably, it further includes a feedback control module, which is used to evaluate the indexes of the processed signal. If it is evaluated that the processed signal does not meet the preset indexes, the feedback control module feeds back a message that the indexes are not reached to the decision-making processing module; According to the message that the indexes are not reached, the decision-making processing module switches to the next clutter suppression method for processing, or increases the polarization dimension information in the clutter suppression method until the feedback control module evaluates that the processed signal meets the preset indexes.

[0013] Preferably, according to the message that the indexes are not reached, switching to the next clutter suppression method for processing, or increasing the polarization dimension information in the clutter suppression method specifically includes: If the type of the corresponding clutter signal corresponds to multiple clutter suppression methods, when receiving the message that the indexes are not reached, switch to the next clutter suppression method for processing until switching to the last clutter suppression method; If after processing the echo signal using the last clutter suppression method, the processed signal still does not meet the preset indexes, select one of the multiple clutter suppression methods and process the echo signal after increasing the polarization dimension information.

[0014] Preferably, it further includes an algorithm library module, which is used to store various clutter suppression methods and various interference suppression methods for the decision-making processing module to select and use.

[0015] Preferably, identifying the type of the clutter signal according to the characteristics of each clutter signal specifically includes: Identifying the type of the clutter signal according to the distance where the clutter signal is located.

[0016] Compared with the prior art, the beneficial effects of the present invention are as follows: By identifying the types of interference signals and clutter signals, and then selecting corresponding suitable processing methods according to the identified types, the present invention can cope with various types of interference and clutter, and improve the detection accuracy of targets in complex environments. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0018] Figure 1 It is a schematic structural diagram of the first cognitive space-time adaptive radar provided by the embodiment of the present invention; Figure 2It is a schematic flowchart of the first method for suppressing unintentional interference of an airborne radar based on frequency-domain adaptive filtering in a cognitive space-time adaptive radar provided by an embodiment of the present invention; Figure 3 It is a schematic flowchart of the second method for suppressing unintentional interference of an airborne radar based on frequency-domain adaptive filtering in a cognitive space-time adaptive radar provided by an embodiment of the present invention; Figure 4 It is a schematic flowchart of the third method for suppressing unintentional interference of an airborne radar based on frequency-domain adaptive filtering in a cognitive space-time adaptive radar provided by an embodiment of the present invention; Figure 5 It is a schematic flowchart of the fourth method for suppressing unintentional interference of an airborne radar based on frequency-domain adaptive filtering in a cognitive space-time adaptive radar provided by an embodiment of the present invention; Figure 6 It is a schematic flowchart of the fifth method for suppressing unintentional interference of an airborne radar based on frequency-domain adaptive filtering in a cognitive space-time adaptive radar provided by an embodiment of the present invention; Figure 7 It is a schematic diagram of the architecture of the second cognitive space-time adaptive radar provided by an embodiment of the present invention; Figure 8 It is a schematic diagram of the architecture of the third cognitive space-time adaptive radar provided by an embodiment of the present invention; Figure 9 It is a schematic flowchart of the sixth method for suppressing unintentional interference of an airborne radar based on frequency-domain adaptive filtering in a cognitive space-time adaptive radar provided by an embodiment of the present invention; Figure 10 It is a schematic diagram of the seventh method for suppressing unintentional interference of an airborne radar based on frequency-domain adaptive filtering in a cognitive space-time adaptive radar provided by an embodiment of the present invention; Figure 11 It is a schematic diagram of the architecture of the fourth cognitive space-time adaptive radar provided by an embodiment of the present invention; Figure 12 It is a schematic diagram of a cognitive space-time adaptive radar provided by an embodiment of the present invention; Figure 13 It is a schematic diagram of another cognitive space-time adaptive radar provided by an embodiment of the present invention; Figure 14 It is a schematic diagram of yet another cognitive space-time adaptive radar provided by an embodiment of the present invention; Figure 15 It is a schematic diagram of still another cognitive space-time adaptive radar provided by an embodiment of the present invention. Detailed implementation manners

[0019] In order to make the objectives, technical solutions and advantages of the present invention more clearly understood, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0020] Unless the context requires otherwise, throughout the specification and claims, the term "comprising" is to be construed in an open inclusive sense, i.e., "including, but not limited to". In the description of the specification, the terms "one embodiment", "some embodiments", "exemplary embodiments", "examples", "specific examples" or "some examples", etc. are intended to indicate that specific features, structures, materials or characteristics related to the embodiment or example are included in at least one embodiment or example of the present disclosure. The schematic representations of the above terms are not necessarily referring to the same embodiment or example. In addition, the specific features, structures, materials or characteristics may be included in any one or more embodiments or examples in any suitable manner, that is, although they may be carried in the embodiments or examples of the above terms due to reasons such as the order of appearance and position, etc., but it is not limited that they can be carried by one embodiment or example in a combined manner.

[0021] In the description of the present invention, the terms "first" and "second" are only used for descriptive purposes and cannot be construed as indicating or implying relative importance or implicitly indicating the quantity of the indicated technical features. Thus, the features defined with "first" and "second" may explicitly or implicitly include one or more of such features. In the description of the embodiments of the present disclosure, unless otherwise specified, the meaning of "a plurality" is two or more. In addition, for example, in the description, for the same type of nouns, the method of adding "A" and "B" at the end is used to describe them as two independent individuals. In this case, the features defined with "A" and "B" are only used for the purpose of distinguishing the same type of individuals and cannot be construed as indicating or implying relative importance or implicitly indicating the quantity of the indicated technical features.

[0022] In the description of the present invention, there will be a description of the form "A and / or B" (where A and B are used to formally represent specific feature contents), and the corresponding description includes the following three combinations: only A, only B, and the combination of A and B.

[0023] As used in the present invention, "about", "substantially" or "approximate" includes the stated value and the average value within an acceptable deviation range of the specific value, where the acceptable deviation range is determined by those of ordinary skill in the art considering the measurement being discussed and the errors associated with the measurement of the specific quantity (i.e., the limitations of the measurement system).

[0024] In addition, the technical features involved in the various embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0025] Example 1: In the prior art, there have been some methods to use prior knowledge such as roads, terrain, and surface coverage types for adaptive signal processing to improve the target detection performance of airborne radars in complex environments. There is also a knowledge-aided STAP (KA-STAP) implementation scheme. With the support of a geographic information database, it effectively suppresses distributed clutter, main-lobe discrete clutter, and side-lobe discrete clutter based on color loading technology, selects uniform samples based on surface coverage type information, and iteratively eliminates interfering targets based on radar tracking information feedback. The effectiveness of this scheme has been verified by the measured data of an X-band 8-channel airborne radar. There is also another implementation scheme, which features mainly color loading technology and supplemented by optimized selection of samples based on prior knowledge. Although the above two schemes give the preliminary implementation process of the KA-STAP technology, they still have the following deficiencies: (1) The refined classification processing of airborne radar clutter has not been achieved; (2) The deep fusion of prior knowledge and radar echo data is insufficient; (3) The problem of interference suppression in complex electromagnetic environments has not been considered.

[0026] To solve the above problems, Embodiment 1 of the present invention provides a cognitive space-time adaptive radar, as Figure 1 shown, including an identification module and a decision-making processing module.

[0027] The identification module is used to identify the type of interference signal using a convolutional neural network and identify the type of clutter signal according to the characteristics of each clutter signal. Among them, those skilled in the art pre-collect data of various interference signals in advance, perform type annotation on the data to obtain a training set, and use the training set to train the original convolutional neural network to obtain the convolutional neural network.

[0028] Identifying the type of clutter signal according to the characteristics of each clutter signal specifically includes: identifying the type of clutter signal according to the distance where the clutter signal is located. For non-stationary clutter, since non-stationarity is generated by clutter blocks closer to the aircraft, the characteristic of non-stationary clutter is that the Doppler characteristic of the clutter is related to the distance. Therefore, the type of clutter signal can be identified through different distances. In actual use, it can be: those skilled in the art preset multiple distance intervals, and one distance interval corresponds to one type of clutter signal. When it is analyzed that the clutter signal is within the corresponding distance interval, it is determined that the type of the clutter signal is the clutter signal type corresponding to that distance interval.

[0029] The decision-making processing module is used to select a corresponding interference suppression method according to the type of interference signal, select a corresponding clutter suppression method according to the type of clutter signal, and process the echo signal using the interference suppression method and the clutter suppression method.

[0030] In this embodiment, by identifying the types of interference signals and clutter signals, and then selecting corresponding suitable processing methods according to the identified types, it is possible to cope with various types of interference and clutter, and improve the detection accuracy of targets in complex environments.

[0031] In a practical application scenario, the types of the clutter signals include one or more of uniform clutter, non-uniform clutter, non-stationary clutter, and wind farm clutter.

[0032] The selecting a corresponding clutter suppression method according to the type of the clutter signal specifically includes: when the type of the clutter signal is uniform clutter, select to use the STAP algorithm (also called the traditional STAP algorithm) or the space-time polarization adaptive processing algorithm for clutter suppression; when the type of the clutter signal is non-uniform clutter, select to use the sample selection method based on geographic information to select uniform training samples, perform power difference correction on the selected uniform training samples, and based on the corrected uniform training samples, select the STAP algorithm or the color loading STAP method for clutter suppression; wherein, the selecting the STAP algorithm or the color loading STAP method for clutter suppression based on the corrected uniform training samples specifically is: estimating the clutter noise covariance matrix using the corrected uniform training samples, then calculating the corresponding filtering vector according to the clutter noise covariance matrix, and processing the clutter signal using the filtering vector. The methods for estimating the clutter noise covariance matrix and calculating the filtering vector are the STAP algorithm or the color loading STAP method.

[0033] When the type of the clutter signal is non-stationary clutter, select to use the 3D-STAP method or the non-stationary clutter suppression method based on space-time clutter spectrum adaptive compensation for clutter suppression; when the type of the clutter signal is wind farm clutter, select to use the wind farm isolated point clutter suppression method based on micro-Doppler characteristics for suppression.

[0034] The space-time polarization adaptive processing algorithm is also called a non-stationary clutter suppression method based on space-time polarization adaptive processing. The geographic information-based sample selection method is also called a prior knowledge-based training sample selection method for post-Doppler STAP. The color loading STAP method is also called an airborne radar knowledge-assisted color loading clutter suppression method and device. The non-stationary clutter suppression method based on space-time clutter spectrum adaptive compensation is also called a conformal array airborne radar space-time clutter spectrum adaptive compensation and clutter suppression method. The wind farm isolated point clutter suppression method based on micro-Doppler features is also called an airborne MIMO radar adaptive clutter suppression method based on space-time sampling matrix.

[0035] In actual use, there is also a situation where wind farm clutter is mixed with other types of clutter signals. At this time, while using the wind farm isolated point clutter suppression method based on micro-Doppler features for interference suppression, other methods corresponding to other types of clutter signals are also used for interference suppression; when one type of clutter signal corresponds to multiple clutter suppression methods, those skilled in the art pre-set the priorities of each method, and in actual use, the method with a higher priority is preferentially selected for processing. For example, in actual use, for uniform clutter, the priority of the STAP algorithm is higher than that of the space-time polarization adaptive processing algorithm; for non-uniform clutter, the priority of the STAP algorithm or is higher than that of the color loading STAP method; for non-stationary clutter, the priority of the 3D-STAP method is higher than that of the non-stationary clutter suppression method based on space-time clutter spectrum adaptive compensation.

[0036] The types of the interference signals include one or more of unintentional interference, main lobe interference, side lobe interference, and main lobe unintentional interference.

[0037] Selecting a corresponding interference suppression method according to the type of the interference signal specifically includes: when the type of the interference signal is unintentional interference or main lobe unintentional interference, selecting and using an airborne radar unintentional interference suppression method based on frequency domain adaptive filtering for interference suppression; when the type of the interference signal is main lobe interference, selecting and using a space-time polarization joint adaptive anti-main lobe interference method for interference suppression; when the type of the interference signal is side lobe interference, selecting and using a STAP method in an interference environment for interference suppression.

[0038] The space-time polarization joint adaptive anti-main lobe interference method is also called an airborne radar space-time polarization joint adaptive processing clutter suppression method. Combining the above various embodiments, the following table can be obtained:

[0039] That is, the decision processing module selects the corresponding clutter suppression method and interference suppression method according to the above table, so as to suppress the clutter signal and the interference signal.

[0040] Among them, the airborne radar unintentional interference suppression method based on frequency-domain adaptive filtering is as follows Figure 2 shown, and specifically includes: In step 201, based on the maximum likelihood criterion, interference spectrum estimation is performed through adaptive beamforming and frequency spectrum estimation algorithms to obtain the power spectral density estimators at each frequency point; among them, this method is based on the subarray fast-time domain sampling data within multiple pulse repetition periods.

[0041] In step 202, according to the power spectral density estimators at each frequency point, the frequency range of the interference signal is determined.

[0042] In step 203, according to the frequency range of the interference signal, multiple channels are constructed, and each channel corresponds to an interference filter.

[0043] In step 204, in each channel, the interference filter is used to filter out the interference signal to obtain the interference-free signal of each channel. For the colored noise caused by filtering, noise compensation is performed on the interference-free signal of each channel to obtain the channel output signal of each channel.

[0044] In step 205, the channel output signal with the minimum power is selected as the final output signal. Among them, the selection of the channel output signal with the minimum power as the final output signal can be to set a power discriminator at the output end of each channel and use the power discriminator to output the channel output signal with the minimum power.

[0045] This method is based on the subarray fast-time domain sampling data within multiple pulse repetition periods, realizes interference spectrum estimation based on the maximum likelihood criterion through adaptive beamforming and frequency spectrum estimation algorithms, and realizes the identification of the frequency range of the interference signal; secondly, for the fast-time sampling signals received by each subarray, interference suppression is performed through frequency-domain filtering; again, the filtered output noise is compensated, thereby completing the suppression of main lobe unintentional interference and improving the target detection performance of the airborne radar.

[0046] Among them, the interference spectrum estimation based on the maximum likelihood criterion through adaptive beamforming and frequency spectrum estimation algorithms to obtain the power spectral density estimators at each frequency point is as follows Figure 3 shown, and specifically includes: In step 301, it is determined that in a Gaussian white noise environment, the output signal corresponding to the u th pulse signal after adaptive beamforming at the corresponding spatial angle is ; among them, is the output signal of the u th pulse interference signal after adaptive beamforming; is the noise signal, and obeys a mean of 0 and a variance of Gaussian distribution.

[0047] In step 302, after performing the mixed-radix FFT transformation on , the observation model at the frequency point is ; where is the observed value of the noise signal intensity at the frequency u in the th observation, which follows a complex Gaussian distribution with a mean of 0 and a variance of , that is ; is the observed value of the interference signal intensity at the frequency u in the th observation; K is the number of groups of received signals.

[0048] In step 303, when the array has received K groups of signals, the estimated power spectral density at the frequency point is calculated as ; where is the estimated value of the amplitude at the frequency point , ; follows a complex Gaussian distribution with a mean of and a variance of , that is ; is the estimated value of the interference signal intensity at the frequency ; is the estimated value of the noise signal intensity at the frequency .

[0049] In a specific application scenario, determining the frequency range of the interference signal based on the estimated power spectral density at each frequency point, as shown in Figure 4 , specifically includes: In step 401, based on the estimated power spectral density of the signal , a signal observation model is established as ; where represents that there is no interference at the frequency point , and represents that there is interference at the frequency point .

[0050] In step 402, based on the signal observation model, the estimated value of the power spectral density without interference at the frequency point is determined; follows an exponential distribution; where the estimated value of the power spectral density without interference is the corresponding estimated value of the power spectral density when there is no interference at the corresponding frequency point.

[0051] In step 403, set the power spectral density estimator of the frequency point as the detection unit, the power spectral density estimator of the frequency point as the protection unit, and the power spectral density estimator of the frequency point as the reference unit. Determine the reference mean according to the power spectral density estimator of the reference unit , and the expression is ; where is the power spectral density estimator at the frequency point , is the power spectral density estimator at the frequency point , is the length of the protection unit, and is the length of the reference unit.

[0052] In step 404, determine whether there is interference at each frequency point according to the power spectral density estimator of each frequency point and the reference mean , and determine the frequency range of the interference signal according to the range of the frequency points with interference .

[0053] Among them, when the power spectral density estimator of the frequency point is , it is determined that there is interference at the frequency point ; when the power spectral density estimator of the frequency point is F , it is determined that there is no interference at the frequency point . It is expressed in the form of a mathematical formula as: .

[0054] In a preferred embodiment, multiple channels are constructed according to the frequency range of the interference signal, and each channel corresponds to an interference filter, as Figure 5 shown. Specifically, it includes: In step 501, according to the frequency range of the interference signal, determine that the reference range is ; where B represents the receiver bandwidth and L represents the number of range cells.

[0055] In step 502, set 2 m +1 channels, and determine the filtering range of the i-th channel as according to the reference range; where i is the channel number, and m is obtained by those skilled in the art through empirical analysis.

[0056] In step 503, based on the all-pass filter, the amplitude-frequency response within the filtering range of the i-th channel is set to zero to obtain the interference filter for the i-th channel, that is, the interference filter for the i-th channel ; where is the bandwidth of the reference range, ; is the center frequency of the filtering range of the i-th channel, ; is a gate function with a width of and a frequency shift of .

[0057] Among them, the input of each channel is the received echo signal, that is, the power of the input signals of each channel is the same, and the output signal of the channel with the minimum power after filtering is the signal with the maximum degree of interference removed.

[0058] In an alternative embodiment, filtering the interference signal using the interference filter in each channel to obtain the interference-free signal of each channel specifically includes: When the echo signal of the corresponding range cell received by the subarray within one pulse repetition period of the airborne radar is , the spectrum of the interference-free signal of the i-th channel obtained by filtering the interference signal using the interference filter of the i-th channel is ; where is the target signal, is the interference signal, is the clutter signal, is the noise signal.

[0059] In a specific application scenario, compensating the noise of the interference-free signal of each channel for the color noise caused by filtering to obtain the channel output signal of each channel, as shown in Figure 6 , specifically includes: In step 601, in the passive mode, calculate the noise power spectral density i of the -th channel through the amplitude values within the interference-free frequency band range; where represents the length of the frequency band not affected by interference.

[0060] In step 602, according to the noise power spectral density i of the -th channel, generate compound Gaussian white noise with a power spectral density of i for the -th channel; among them, the compound Gaussian white noise with a power spectral density of is generated by the computer in the way of generating random numbers obeying the Gaussian distribution.

[0061] In step 603, a noise filter is constructed for the i-th channel ; wherein ; the compound Gaussian white noise is passed through the noise filter of the i-th channel and then superimposed on the output channel of the interference filter of the i-th channel to achieve noise compensation for the interference-free signal.

[0062] Using the above airborne radar unintentional interference suppression method based on frequency-domain adaptive filtering can effectively reduce the influence brought by the change of the interference signal frequency and the estimation error, so as to effectively suppress the unintentional interference or the main lobe unintentional interference.

[0063] In a preferred embodiment, as Figure 7 shown, the cognitive space-time adaptive radar described in this embodiment further includes a feedback control module, and the feedback control module is used to evaluate the index of the processed signal. If it is evaluated that the processed signal does not meet the preset index, a message that the index is not reached is fed back to the decision-making processing module; The decision-making processing module switches to the next clutter suppression method for processing according to the message that the index is not reached, or increases the polarization dimension information in the clutter suppression method until the feedback control module evaluates that the processed signal meets the preset index. The preset index is obtained by those skilled in the art through empirical analysis. In actual use, the preset index may be that the Minimum Detectable Velocity (MDV) is less than a preset value, and the preset value is obtained by those skilled in the art through requirement analysis.

[0064] Wherein, the switching to the next clutter suppression method for processing according to the message that the index is not reached, or increasing the polarization dimension information in the clutter suppression method specifically includes: If there are multiple clutter suppression methods corresponding to the type of the corresponding clutter signal, when receiving the message that the index is not reached, switch to the next clutter suppression method for processing until switching to the last clutter suppression method; If the processed signal still does not meet the preset criteria after processing the echo signal using the last clutter suppression method, one of the multiple clutter suppression methods is selected to process the echo signal after adding polarization dimension information. For example, if the processed signal fails to meet the preset criteria after processing homogeneous clutter using the traditional STAP algorithm, switch to using the space-time polarization adaptive processing algorithm to process the homogeneous clutter. If the preset criteria still cannot be met after processing homogeneous clutter using the space-time polarization adaptive processing algorithm, any one of the traditional STAP algorithm or the space-time polarization adaptive processing algorithm can be used to process the homogeneous clutter after adding polarization dimension information. In actual use, a method with a processed signal relatively closer to the preset criteria can be selected to add polarization dimension information and then process.

[0065] In a specific application scenario, such as Figure 8 shown, it further includes an algorithm library module, which is used to store various clutter suppression methods and various interference suppression methods for the decision-making processing module to select and use.

[0066] Example 2: Based on the method described in Embodiment 1, the present invention combines a specific application scenario and elaborates on the airborne radar unintentional interference suppression method based on frequency-domain adaptive filtering in Embodiment 1 through technical expressions in the relevant scenario.

[0067] The airborne radar unintentional interference suppression method based on frequency-domain adaptive filtering provided in this embodiment specifically includes: Based on the subarray fast-time domain sampling data within multiple pulse repetition periods, interference spectrum estimation is achieved through adaptive beamforming and frequency spectrum estimation algorithms based on the maximum likelihood criterion. The Constant False Alarm Rate Detector (CFAR) is used to determine the threshold to identify the frequency range of the interference signal; in the passive mode, the noise power spectral density is estimated using the power spectral density within the interference-free frequency band range; for the fast-time sampling signal received by each subarray, the amplitude-frequency response within the interference signal frequency range is set to zero; for the colored noise caused by filtering, a noise compensation link is set so that the output noise still follows a Gaussian distribution; in order to reduce the influence brought by the frequency change of the interference signal and the estimation error, based on the estimated interference signal frequency range as a reference, 2 m +1 groups of channels composed of band-stop filters h 1,i and band-pass filters h 2,i are configured, and their stopband and passband ranges are respectively . Finally, a power comparator is set at the output end of each channel, and the signal of the channel with the lowest output power is selected as the final output signal. Here, B represents the receiver bandwidth and L represents the number of range cells. Specifically, as shown in Figure 9 and Figure 10 shown below, it includes the following steps: In step 701, the frequency range of the interference signal is identified, specifically including: in a Gaussian white noise environment, the output signal corresponding to a certain spatial angle after the first pulse signal passes through adaptive beamforming is:

[0068] where is the output signal of the interference signal of the first pulse after adaptive beamforming, is the noise signal, and the noise signal follows a Gaussian distribution with a mean of 0 and a variance of .

[0069] After performing a mixed-radix FFT transformation on this signal, we get .

[0070] Since the interference pattern is narrowband interference, the interference signal only appears as peaks at finite frequency points in the frequency domain, while the noise is Gaussian white noise, and it still follows a Gaussian distribution at the frequency points in the frequency domain. Therefore, when the array receives K groups of signals, that is, observes the signals K times, the observation model at the frequency point is:

[0071] where is the observed value of the noise signal intensity at the frequency u in the th observation. It follows a complex Gaussian distribution with a mean of 0 and a variance of , that is , is the observed value at the frequency in the th observation.

[0072] According to the maximum likelihood estimation, the estimator of the signal at the frequency point is:

[0073] where follows a complex Gaussian distribution with a mean of and a variance of , that is , is the estimated value of the interference signal intensity at the frequency , is the estimated value at the frequency The estimated value of the noise signal intensity at the location.

[0074] Then the power spectral density of the signal is:

[0075] Based on the power spectral density estimation model, the above interference frequency band estimation problem can be transformed into a typical binary hypothesis testing problem, and its observation model can be expressed as:

[0076] Since when there is no interference signal at the frequency point the estimated value of its power spectral density is:

[0077] Then obeys an exponential distribution. According to the CFAR principle, set the estimated value of the power spectral density at the frequency point as the detection unit, the estimated value of the power spectral density at the frequency point as the protection unit, and the estimated value of the power spectral density at the frequency point as the reference unit. According to the estimated value of the power spectral density of the reference unit, determine the reference mean with the expression where is the estimated value of the power spectral density at the frequency point , is the estimated value of the power spectral density at the frequency point , is the length of the protection unit, is the length of the reference unit; after determining the false alarm probability P F according to historical experience, determine whether there is interference in the detection unit through the following formula:

[0078] where .

[0079] Detect each frequency point respectively to obtain the frequency range of the interference signal. To reduce the estimation error caused by the spectral resolution, on the basis of identifying the frequency range of the interference signal, further set the interference frequency range as:

[0080] where B represents the receiver bandwidth and L represents the number of range cells.

[0081] In step 702, perform noise power spectral density estimation, specifically including: in the passive mode, estimate the noise power spectral density of the channel by using the power spectral density in the interference-free frequency band range of the i th channel, that is:

[0082] wherein represents the length of the frequency band not affected by interference.

[0083] In step 703, frequency-domain filtering is performed, specifically including: based on the estimated interference frequency range, on the basis of an all-pass filter, setting the amplitude-frequency response within the interference signal frequency range to zero. In the active mode, assuming that the echo signal of a certain range cell received by the subarray of the airborne radar within a pulse repetition period is:

[0084] wherein is the target signal, is the interference signal, is the clutter signal, is the noise signal.

[0085] Construct the interference filter for the i-th channel , by estimating the frequency range of the interference signal , making the response of the interference filter zero in this frequency range, that is:

[0086] wherein is the filter bandwidth; is the center frequency of the interference signal; is a rectangular function with a width of and a frequency shift of .

[0087] Therefore, the output signal spectrum of the i-th channel after frequency-domain filtering is:

[0088] In step 704, noise compensation is performed, specifically including: in order to eliminate the influence of colored noise, a noise compensation link is added, specifically including: Since in the passive mode, the noise power spectral density of the i-th channel has been estimated through the amplitude values within the interference-free frequency band range, that is ; wherein, represents the length of the frequency band not affected by interference.

[0089] According to the estimated power spectral density , for the i-th channel, corresponding Gaussian white noise is simulated and generated within the system.

[0090] Construct the noise filter , and its amplitude-frequency response and the interference filter have the following relationship:

[0091] Pass the Gaussian white noise signal obtained by simulation through a noise filter and then superimpose it on the output channel of the interference filter .

[0092] In step 705, multi-channel decision is performed to obtain the output signal, specifically including: setting a power comparator at the output end of each channel, and selecting the signal of the channel with the lowest output power as the final output signal.

[0093] In this embodiment, a method for suppressing the main lobe unintentional interference of an airborne radar based on frequency domain filtering is provided. First, based on the subarray fast time domain sampling data within multiple pulse repetition periods, interference spectrum estimation is realized through adaptive beamforming and frequency spectrum estimation algorithms based on the maximum likelihood criterion, and the frequency range of the interference signal is identified by using CFAR to determine the threshold; secondly, for the fast time sampling signals received by each subarray, interference suppression is performed through frequency domain filtering; thirdly, the filtered output noise is compensated; finally, multi-channel decision is performed to complete the suppression of the main lobe unintentional interference, improving the target detection performance of the airborne radar.

[0094] This embodiment can realize the accurate estimation of interference parameters; moreover, the fast time frequency domain filtering technology proposed in this embodiment realizes the effective suppression of narrowband interference signals with different modulation methods without changing the length of the fast time sampling data; and, through noise compensation processing, the influence of colored noise on the target detection performance after filtering processing is effectively solved.

[0095] Example 3: This embodiment also conducts a more detailed functional division of the cognitive space-time adaptive radar from another perspective, such as Figure 11 shown, including a database module, an algorithm library module, an identification module, a decision processing module, and a feedback control module.

[0096] Among them, the database module is used to store data in four parts, namely geographic information data, radar data, platform system parameters, and radar system parameters, providing important theoretical support for subsequent airborne radar zoning processing. Among them, the geographic information database mainly contains two types of data, namely surface coverage data and digital elevation model data; the radar database contains measured data of multiple types of airborne radars in different terrain environments and electromagnetic environments. Among them, the terrain environments include cities, mountains, deserts, hills, seas, land-sea boundaries, etc.; the interference patterns include suppression noise interference, deception interference, smart interference, and unintentional communication interference, etc.; the platform system parameters include parameters such as the flight altitude, speed, and heading of the airborne radar; the radar system parameters include radar wavelength, beam pointing, beam width, pulse repetition frequency, and array offset angle, etc.

[0097] The algorithm library module includes an expert decision-making unit, a model library unit, and an STAP algorithm library unit. Among them, the STAP algorithm library unit is used to store airborne clutter suppression algorithm models, which contain more than a hundred airborne radar STAP algorithms to cope with dynamic terrain and complex electromagnetic environments. For example: reduced-dimension STAP, reduced-rank STAP, STAP in interference environment, non-uniform STAP, non-stationary STAP, STAP with new system, and STAP angle measurement, etc. The model library unit is based on prior knowledge such as the surface cover database and the digital elevation model database, and according to the refined spatio-temporal clutter signal model of the airborne radar, it realizes the simulation of the echo of the airborne radar. This unit first maps many information points contained in the geographic database to each clutter block, secondly obtains the clutter scattering coefficients of the information points based on the statistical analysis model of the Lincoln Laboratory experimental data, and calculates the resolution cell area of the information points. Thirdly, it conducts information point occlusion judgment, and finally obtains the refined clutter signal of the airborne radar. Based on this simulation method, it is possible to achieve high-fidelity simulation of the echo of the airborne radar under complex terrain environments, complex electromagnetic environments, and complex target environments. At the same time, it is possible to simulate the echoes of conventional airborne radars, conformal array airborne radars, end-fire array airborne radars, bistatic airborne radars, and MIMO airborne radars. The expert decision-making unit selects the STAP algorithm with the highest matching degree with each clutter area and interference based on certain rules according to the distribution characteristics and interference patterns of the airborne radar clutter in the range-Doppler two-dimensional domain for filtering processing to achieve the global optimum of the target detection performance of the airborne radar. When there are two algorithms corresponding to the same echo type, the first algorithm has a higher priority. If the index requirements are not met, the second algorithm is selected for processing. The decision rules stored in the expert decision-making unit are shown in the following table:

[0098] The recognition module first realizes interference recognition based on a convolutional neural network; secondly, according to the clutter signals received by the airborne radar in the actual working environment, the clutter signals are classified into uniform clutter, non-uniform clutter, non-stationary clutter, wind farm clutter, main lobe clutter, etc.; finally, according to the clutter characteristics of different regions, a specific STAP algorithm is selected for clutter suppression. Among them, the convolutional neural network does not require an additional feature extraction process, and only needs to input the labeled sample data into the network to automatically extract image features using the convolutional layer.

[0099] After recognizing various types of clutter signals and interference signals, the decision processing module selects the corresponding processing methods from the algorithm library module for processing, and the various processing methods are shown in the above table.

[0100] Among them, the suppression process of non-uniform clutter is as Figure 12As shown in the figure, the method first constructs a prior knowledge information database using airborne radar platform parameters, topographic and geomorphic parameters, and geographic elevation parameters; secondly, it determines the information points corresponding to each range-Doppler cell; thirdly, it constructs a weighted Euclidean distance measure based on landform type and landform slope, as shown in Equations (1) and (2), to complete the selection of uniform samples, that is, to select uniform samples using a sample selection method based on geographic information; finally, it corrects the power difference of the selected samples, and then uses the corrected uniform training samples for clutter suppression.

[0101] Assume that the k-th Doppler channel and the l-th range-Doppler (RD) cell are the cells to be detected. Then the normalized vectors of the landform types of this cell and its two adjacent RD cells on both sides are respectively: (1) Where represents the proportion of 8 landform types in the l-th RD cell.

[0102] The weighted Euclidean distance measure is: (2) Where α, β, and γ represent the weighting coefficients.

[0103] The process of suppressing non-stationary clutter using a non-stationary clutter suppression method based on spatio-temporal clutter spectrum adaptive compensation is as Figure 13 shown. This method first selects uniform samples based on geographic information, and transforms the conformal array receiving subarray into an equivalent uniform linear array through virtual uniform linear array transformation; secondly, it transforms the spatio-temporal echo data into the spatio-temporal four-dimensional power spectrum domain, as shown in Equation (3), to realize the decoupling of non-stationary clutter and targets; thirdly, on the basis of the selected uniform samples, it compensates the short-range non-stationary clutter to the maximum unambiguous range cell; finally, it uses the compensated spatio-temporal data and the long-range stationary clutter data to estimate the clutter covariance matrix of the cell to be detected, forms spatio-temporal adaptive weights, and completes the clutter suppression process.

[0104] (3) Where represents the spatio-temporal four-dimensional steering vector, and R l represents the echo covariance matrix.

[0105] The process of wind farm clutter suppression is as Figure 14 shown. This method first processes the echo range-Doppler spectrogram using the Sobel operator, and combines geographic information to judge the range gate where the wind turbine is located; secondly, it constructs dictionary atoms using the estimated micro-Doppler characteristic parameters such as the rotational speed and initial phase of the wind turbine blades; finally, it uses the orthogonal projection matrix formed by the dictionary atoms to suppress the wind farm clutter.

[0106] The process of the airborne radar unintentional interference suppression method based on frequency-domain adaptive filtering is as follows Figure 15 As shown, this method first cognizes the external electromagnetic environment through adaptive beamforming (Adaptive Digital Beam Forming, abbreviated as: ADBF) technology in the passive mode to obtain the interference signals at each wave position, then estimates the frequency and bandwidth of the interference signals based on the maximum likelihood criterion and the CFAR criterion, and finally suppresses the main lobe unintentional interference through fast frequency-domain filtering in the active mode. At the same time, the colored noise after filtering is whitened into white noise through noise compensation, which further improves the target detection performance while achieving interference suppression.

[0107] The feedback control module determines whether it is necessary to add polarization dimension information to improve the radar performance by judging whether the performance of the airborne radar after anti-interference and clutter suppression processing meets the index requirements. For example, it judges whether the performance of the airborne radar after anti-interference and clutter suppression processing meets the technical and tactical index requirements through the result of CFAR detection, or measures it through the minimum detectable speed index. If this index cannot meet the requirements, the polarization dimension information is added in the clutter suppression link to improve the radar performance.

[0108] It should be noted here that the processing methods in Embodiment 1 and the method in Embodiment 2 are all applicable in this embodiment and will not be elaborated here.

[0109] It is worth noting that the information interaction, execution process, etc. between the modules and units in the above-mentioned device and system, due to being based on the same concept as the method embodiment of the present invention, the specific content can be referred to the description in the method embodiment of the present invention and will not be elaborated here.

[0110] Those of ordinary skill in the art can understand that all or part of the steps in the various methods of the embodiments can be completed by instructing relevant hardware through a program, and this program can be stored in a computer-readable storage medium. The storage medium can include: read-only memory (ROM, Read Only Memory), random access memory (RAM, Random Access Memory), magnetic disk or optical disc, etc.

[0111] The above are only the preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent replacements, and improvements made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A cognitive space-time adaptive radar, characterized in that: It includes a recognition module and a decision processing module; The identification module is used to identify the type of interference signal using a convolutional neural network, and to identify the type of clutter signal according to the characteristics of each clutter signal; The decision processing module is used to select a corresponding interference suppression method according to the type of interference signal, and select a corresponding clutter suppression method according to the type of clutter signal, and use the interference suppression method and the clutter suppression method to process the echo signal.

2. The cognitive space-time adaptive radar according to claim 1, characterized in that: The types of the clutter signals include one or more of uniform clutter, non-uniform clutter, non-stationary clutter and wind farm clutter; The method of selecting a corresponding clutter suppression method according to the type of clutter signal specifically includes: When the type of the clutter signal is uniform clutter, selecting to use a STAP algorithm or a space-time polarization adaptive processing algorithm to suppress clutter; When the type of the clutter signal is non-uniform clutter, a sample selection method based on geographic information is selected to select uniform training samples, power difference correction is performed on the selected uniform training samples, and based on the corrected uniform training samples, a STAP algorithm or a color-loaded STAP method is selected to suppress clutter; When the type of the clutter signal is non-stationary clutter, a 3D-STAP method or a non-stationary clutter suppression method based on space-time clutter spectrum adaptive compensation is selected to perform clutter suppression; When the type of the clutter signal is wind farm clutter, a wind farm isolated point clutter suppression method based on micro-Doppler characteristics is selected for suppression.

3. The cognitive space-time adaptive radar according to claim 1, characterized in that: The type of the interference signal includes one or more of unintentional interference, main lobe interference, side lobe interference and main lobe unintentional interference; The selecting a corresponding interference suppression method according to the type of the interference signal specifically includes: When the type of the interference signal is unintentional interference or main lobe unintentional interference, selecting to use an airborne radar unintentional interference suppression method based on frequency domain adaptive filtering to perform interference suppression; When the type of the interference signal is main lobe interference, a space-time polarization combined adaptive anti-main lobe interference method is selected to perform interference suppression; When the type of the interference signal is sidelobe interference, the STAP method under the interference environment is selected to perform interference suppression.

4. The cognitive space-time adaptive radar according to claim 3, characterized in that: The airborne radar unintentional interference suppression method based on frequency domain adaptive filtering specifically includes: Based on the maximum likelihood criterion, the interference spectrum is estimated through adaptive beamforming and frequency spectrum estimation algorithm to obtain the power spectrum density estimate of each frequency point; Determine the frequency range of the interference signal based on the power spectrum density estimation of each frequency point; According to the frequency range of the interference signal, a plurality of channels are constructed, each channel corresponding to an interference filter; In each channel, an interference filter is used to filter out the interference signal to obtain the interference-free signal of each channel. The interference-free signal of each channel is compensated for the color noise caused by filtering to obtain the channel output signal of each channel. The channel output signal with the smallest power is selected as the final output signal.

5. The cognitive space-time adaptive radar according to claim 4, characterized in that: The interference spectrum estimation is performed based on the maximum likelihood criterion through adaptive beamforming and frequency spectrum estimation algorithm to obtain the power spectrum density estimation of each frequency point, specifically including: Determine in a Gaussian white noise environment, u After the pulse signal is adaptively beamformed, the output signal corresponding to the corresponding spatial angle is ;in, For the u The output signal of the interference signal of a pulse after adaptive beamforming; is a noise signal, and The mean is 0 and the variance is Gaussian distribution of right After mixed basis FFT transformation, we get The observation model at is ;in, For the u In the observations, the frequency The observed value of the noise signal strength at , which has a mean of 0 and a variance of The complex Gaussian distribution of ; For the u In the observations at frequency The observed value of the interference signal strength at K is the number of groups of received signals; When the array receives K When the group signal is calculated, the frequency The power spectral density estimator at is ;in, For the frequency The amplitude estimate at , ; The mean is , the variance is The complex Gaussian distribution of ; For the frequency An estimate of the interference signal strength at For the frequency An estimate of the noise signal strength at .

6. The cognitive space-time adaptive radar according to claim 5, characterized in that: Determining the frequency range of the interference signal according to the power spectrum density estimate of each frequency point specifically includes: According to the power spectral density estimate of the signal , the signal observation model is established as ;in, Representative frequency No interference, Representative frequency There is interference; According to the signal observation model, the frequency point is determined The interference-free power spectral density estimator ; Obeying exponential distribution; wherein the interference-free power spectrum density estimator is the power spectrum density estimator corresponding to the absence of interference at the corresponding frequency point; Set frequency The power spectral density estimator is the detection unit, frequency point The power spectral density estimator is the protection unit, the frequency point The power spectral density estimator of the reference unit is used as the reference unit. According to the power spectral density estimator of the reference unit, the reference mean is determined , the expression is ;in, For the frequency The power spectral density estimator of For the frequency The power spectral density estimator of is the length of the protection unit, is the length of the reference unit; According to the power spectrum density estimate of each frequency point and the reference mean , determine whether there is interference at each frequency point, and determine the frequency range of the interference signal based on the range of the frequency points where interference exists ; Among them, when the frequency The power spectral density estimator When the frequency is determined There is interference; when the frequency The power spectral density estimator When the frequency is determined There is no interference; where P F is the false alarm probability determined based on historical experience, .

7. The cognitive space-time adaptive radar according to claim 1, characterized in that: It also includes a feedback control module, which is used to perform an index evaluation on the processed signal, and if the evaluation shows that the processed signal does not meet the preset index, it will feedback a message that the index is not met to the decision processing module; The decision processing module switches to the next clutter suppression method for processing according to the indicator not being met message, or adds polarization dimension information in the clutter suppression method, until the feedback control module evaluates that the processed signal meets the preset indicator.

8. The cognitive space-time adaptive radar according to claim 7, characterized in that: The step of switching to the next clutter suppression method for processing according to the indicator not being reached message, or adding polarization dimension information in the clutter suppression method, specifically includes: If the type of the corresponding clutter signal corresponds to multiple clutter suppression methods, when receiving a message that the indicator is not reached, switching to the next clutter suppression method for processing until switching to the last clutter suppression method; If the echo signal still does not meet the preset indicator after being processed by the last clutter suppression method, one of the multiple clutter suppression methods is selected to process the echo signal after adding polarization dimension information.

9. The cognitive space-time adaptive radar according to claim 1, characterized in that: It also includes an algorithm library module, which is used to store various clutter suppression methods and various interference suppression methods for selection and use by the decision processing module.

10. The cognitive space-time adaptive radar according to claim 1, characterized in that: The method of identifying the type of clutter signals according to the characteristics of each clutter signal specifically includes: Identify the type of clutter signal based on the distance at which the clutter signal is located.

Images