Sequentially optimized isar multi-narrowband jamming detection and suppression method

Abstract

The application discloses a kind of sequential optimization ISAR multi narrowband interference detection and suppression method, it is related to radar signal processing technical field, including: obtaining echo signal, constructs radar range imaging model under narrowband interference environment;Based on radar range imaging model, construct anti narrowband interference waveform optimization model;The spectrum of echo signal is divided into several equal-length sub-frequency bands, the amplitude modulation coefficient of each sub-frequency band is set to the same value, the amplitude modulation coefficient of each sub-frequency band is solved using GA algorithm, and the sub-frequency band where narrowband interference is located is roughly estimated;The proportion of each sub-frequency band in the spectrum is solved using PSO algorithm, and the frequency band where narrowband interference is located is precisely estimated;Based on the estimation result of the frequency band where narrowband interference is located, a spectral notch filter is constructed, and interference suppression is achieved after matched filtering.The present application can obtain accurate and optimized waveform power spectrum estimation result under multi narrowband interference environment.

Description

Technical Field

[0001] This invention belongs to the field of radar signal processing technology, specifically relating to a sequentially optimized ISAR multi-narrowband interference detection and suppression method. Background Technology

[0002] Inverse Synthetic Aperture Radar (ISAR) imaging enables all-weather, all-day, long-range target observation and high-resolution imaging, finding wide application in situational awareness, air defense, and missile defense. However, the human-occupied space is filled with various electromagnetic devices, inevitably subjecting radar systems to intentional or unintentional electromagnetic interference in the same frequency band. Compared to the broadband signal emitted by the ISAR system, the interference signal is usually single-frequency or narrowband interference. This interference signal leads to a decrease in the quality of ISAR echo range imaging, making it impossible to achieve focused imaging of the observed target. Therefore, adaptive detection and suppression of narrowband interference in broadband echo signals has become a hot topic in the field of radar anti-jamming waveform optimization.

[0003] In related technologies, narrowband interference suppression methods based on signal power spectrum optimization are mainly divided into two categories. The first category is anti-interference waveform optimization methods based on different criteria, such as the maximum signal-to-noise ratio criterion and the maximum mutual information criterion. These methods have difficult-to-determine parameters and require prior knowledge of interference and the target. The second category is anti-interference waveform optimization methods based on sparse spectrum. This category can be further subdivided into two types: the first is anti-interference waveform optimization methods based on intelligent optimization algorithms, such as the most common Particle Swarm Optimization (PSO) algorithm. These methods typically select an effective index representing the optimization quality as the objective function, and adjust parameters such as signal phase and amplitude sequence through intelligent optimization algorithms to obtain a signal that meets the conditions as the optimized waveform output. The advantage of this type of algorithm is its simple structure and strong applicability; the disadvantage is that the result is highly dependent on the objective function, and current anti-interference waveform optimization methods based on this type of algorithm cannot work independently of prior knowledge of interference. The second type is waveform optimization methods based on sparse spectrum construction using alternating projection, with a typical example being the Stopband Cyclic Algorithm New... The methods include the New SCAN algorithm and the weighted stopband cyclic algorithm (Cyclic Algorithm New, WeSCAN). Although these methods take into account both signal sidelobe performance and anti-interference performance, they require manual setting of the weighting coefficient between the two. The algorithm performance depends on manual parameter tuning and requires prior knowledge of the interference.

[0004] Currently, most anti-interference waveform optimization methods based on signal power spectrum require prior knowledge of the interference, or even prior knowledge of the target. However, in actual ISAR imaging scenarios, there is not only friendly electromagnetic interference but also enemy electromagnetic interference. Furthermore, the ISAR observation targets are usually non-cooperative targets, and it is impossible to predict the relevant characteristics of the interference signal and the target. In this case, the above methods have significant limitations. Therefore, there is an urgent need for a multi-narrowband interference detection and suppression method. Summary of the Invention

[0005] To address the aforementioned problems in the existing technology, this invention provides a sequentially optimized ISAR multi-narrowband interference detection and suppression method. The technical problem to be solved by this invention is achieved through the following technical solution:

[0006] In a first aspect, this application provides a sequentially optimized ISAR multi-narrowband interference detection and suppression method, including:

[0007] Step S101: Acquire echo signals and construct a radar range imaging model under narrowband interference environment;

[0008] Step S102: Based on the radar range imaging model, obtain the spectrum of the echo signal and construct the amplitude modulation coefficient vector of the echo signal spectrum. Further, construct an anti-narrowband interference waveform optimization model; wherein, the anti-narrowband interference waveform optimization model is related to the amplitude modulation coefficient vector of the echo signal spectrum.

[0009] Step S103: Divide the spectrum of the echo signal into several sub-bands of equal length, set the amplitude modulation coefficient of each sub-band to the same value, and use the GA algorithm to solve the amplitude modulation coefficient of each sub-band to make a coarse estimate of the sub-band where the narrowband interference is located; wherein, the amplitude modulation coefficient vector of the spectrum of the echo signal includes multiple amplitude modulation coefficients, and each sub-band corresponds to one amplitude modulation coefficient.

[0010] Step S104: Use the PSO algorithm to solve for the proportion of each sub-band in the spectrum of the echo signal, and make a precise estimate of the frequency band where the narrowband interference is located; where the frequency band is the frequency range of the narrowband interference in the spectrum of the echo signal.

[0011] Step S105: Construct a spectrum notch filter based on the estimation results of the frequency band where the narrowband interference is located, and achieve interference suppression after matched filtering.

[0012] The beneficial effects of this invention are:

[0013] This invention provides a sequentially optimized ISAR multi-narrowband interference detection and suppression method. The ISAR system receives the echo signal and transforms the anti-interference waveform optimization problem into the problem of optimizing the amplitude modulation coefficient vector of the reference signal spectrum. To address the problem of high dimensionality and difficulty in convergence during amplitude modulation coefficient vector optimization, the GA algorithm is used to obtain the number of narrowband interferences, and the PSO algorithm is used to obtain the frequency band of the narrowband interferences. Then, a notch filter is constructed to effectively suppress narrowband interferences and optimize the waveform, thereby achieving target focusing imaging. This application overcomes the problem of existing algorithms relying on prior knowledge of interference and targets, and does not require a complex parameter tuning process. It can obtain accurate optimized waveform power spectrum estimation results in multi-narrowband interference environments.

[0014] The present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0015] Figure 1 This is a flowchart of a sequentially optimized ISAR multi-narrowband interference detection and suppression method provided in an embodiment of the present invention;

[0016] Figure 2 This is a flowchart of a multi-narrowband interference detection algorithm based on the GA-PSO sequential optimization algorithm provided in an embodiment of the present invention;

[0017] Figure 3 This is a schematic diagram of a two-dimensional point target model of an aircraft provided in an embodiment of the present invention;

[0018] Figure 4(a) is a schematic diagram showing the comparison results of the optimized waveform and the interference power spectrum provided in the embodiment of the present invention;

[0019] Figure 4(b) is a schematic diagram comparing the one-dimensional range profile results of isolated scattering points of the optimized waveform and the unoptimized waveform provided in the embodiment of the present invention;

[0020] Figure 4(c) is a result diagram of two-dimensional imaging of unoptimized waveform and optimized waveform provided in an embodiment of the present invention;

[0021] Figure 4(d) is another result of two-dimensional imaging of unoptimized and optimized waveforms provided in the embodiment of the present invention. Detailed Implementation

[0022] The present invention will be further described in detail below with reference to specific embodiments, but the implementation of the present invention is not limited thereto.

[0023] Please see Figure 1 , Figure 1 This is a flowchart of a sequentially optimized ISAR multi-narrowband interference detection and suppression method provided in this embodiment of the invention. Please refer to it. Figure 1As shown, the sequentially optimized ISAR multi-narrowband interference detection and suppression method provided in this application includes:

[0024] Step S101: Acquire echo signals and construct a radar range imaging model under narrowband interference environment;

[0025] Step S102: Based on the radar range imaging model, obtain the spectrum of the echo signal and construct the amplitude modulation coefficient vector of the echo signal spectrum. Further, construct an anti-narrowband interference waveform optimization model; wherein, the anti-narrowband interference waveform optimization model is related to the amplitude modulation coefficient vector of the echo signal spectrum.

[0026] Step S103: Divide the spectrum of the echo signal into several sub-bands of equal length, set the amplitude modulation coefficient of each sub-band to the same value, and use the GA algorithm to solve the amplitude modulation coefficient of each sub-band to make a coarse estimate of the sub-band where the narrowband interference is located; wherein, the amplitude modulation coefficient vector of the spectrum of the echo signal includes multiple amplitude modulation coefficients, and each sub-band corresponds to one amplitude modulation coefficient.

[0027] Step S104: Use the PSO algorithm to solve for the proportion of each sub-band in the spectrum of the echo signal, so as to achieve a precise estimation of the frequency band where the narrowband interference is located; where the frequency band is the frequency range of the narrowband interference in the spectrum of the echo signal.

[0028] Step S105: Construct a spectrum notch filter based on the estimation results of the frequency band where the narrowband interference is located, and achieve interference suppression after matched filtering.

[0029] For details, please refer to [link / reference]. Figure 1 As shown in this embodiment, the sequentially optimized ISAR multi-narrowband interference detection and suppression method can accurately estimate the position of the narrowband interference spectrum in the echo signal after the Inverse Synthetic Aperture Radar (ISAR) system receives the echo signal, and can optimize the reference signal spectrum to achieve effective suppression of narrowband interference; the process is described in detail below.

[0030] Step S101: Receive the echo signal through ISAR and construct a radar range imaging model under narrowband interference environment based on the echo signal;

[0031] Step S102: Based on the radar range imaging model constructed in step S101 above, obtain the spectrum of the echo signal, and construct the amplitude modulation coefficient vector of the echo signal spectrum. Further, construct an anti-narrowband interference waveform optimization model; wherein, the anti-narrowband interference waveform optimization model is related to the amplitude modulation coefficient vector of the echo signal spectrum.

[0032] Step S103: Divide the spectrum of the echo signal into several sub-bands of equal length, set the amplitude modulation coefficient of each sub-band to the same value, and use the GA algorithm to solve the amplitude modulation coefficient of each sub-band to make a coarse estimate of the sub-band where the narrowband interference is located; wherein, the amplitude modulation coefficient vector of the spectrum of the echo signal includes multiple amplitude modulation coefficients, and each sub-band corresponds to one amplitude modulation coefficient.

[0033] Step S104: The PSO algorithm is used to solve for the proportion of each sub-band in the spectrum of the echo signal, so as to achieve a precise estimation of the frequency band where the narrowband interference is located; where the frequency band is the frequency range of the narrowband interference in the spectrum of the echo signal, and the sub-band is the spectrum division of the echo signal obtained by this method to solve for the frequency band where the narrowband interference is located.

[0034] Step S105: Construct a spectrum notch filter based on the estimation results of the frequency band where the narrowband interference is located, and achieve interference suppression after matched filtering.

[0035] Through the above steps, the ISAR system receives the echo signal and transforms the anti-interference waveform optimization problem into the problem of optimizing the amplitude modulation coefficient vector of the reference signal spectrum. To address the problem of high dimensionality and difficulty in convergence during amplitude modulation coefficient vector optimization, the GA algorithm is used to obtain the number of narrowband interferences, and the PSO algorithm is used to obtain the frequency band where the narrowband interferences are located. Then, a notch filter is constructed to effectively suppress narrowband interferences and optimize the waveform, thereby achieving target focusing imaging. This application overcomes the problem of existing algorithms relying on prior knowledge of interference and targets, and does not require a complex parameter tuning process. It can obtain accurate optimized waveform power spectrum estimation results in multi-narrowband interference environments.

[0036] In one optional embodiment of this application, the specific process of acquiring the echo signal and constructing the radar range imaging model under narrowband interference environment in step S101 is as follows:

[0037] Step S1011: Obtain the noise narrowband interference signal ;

[0038] Step S1012: Receive echo signal , The echo signal contains noisy narrowband interference signals, among which, To observe the noise, To observe the echo signal of the target, L represents the number of equivalent scattering centers of the echo signal from the observed target. For the first Backscattering coefficient of each scattering center For distance window, The width of the radar-transmitted pulse signal is given by [insert value here]. , The first echo signal of the observed target The instantaneous distance between the scattering center and the radar. The speed of electromagnetic wave propagation. The carrier frequency of the pulse signal transmitted by the radar. The frequency modulation frequency of the broadband signal transmitted by the radar;

[0039] Step S1013: Perform pulse compression processing based on the echo signal to obtain the radar range imaging model under narrowband interference environment; the radar range imaging model is as follows: ,in, For reference signal, This is for convolution processing.

[0040] Specifically, in this embodiment, a radar range imaging model under narrowband interference environment is constructed through the following steps:

[0041] Step S1011: Obtain narrowband interference signal ;

[0042] Step S1012: The ISAR system receives the echo signal. Its expression is The echo signal contains noisy narrowband interference signals, among which, To observe the noise, The echo signal of the observed target is expressed as follows: The echo signal from the observed target includes information such as the target's electromagnetic scattering characteristics and motion characteristics; L represents the number of equivalent scattering centers of the echo signal from the observed target. For the first Backscattering coefficient of each scattering center For distance window, The width of the radar-transmitted pulse signal is given by [insert value here]. , The first echo signal of the observed target The instantaneous distance between the scattering center and the radar. The speed of electromagnetic wave propagation. The carrier frequency of the pulse signal transmitted by the radar. The frequency modulation frequency of the broadband signal transmitted by the radar;

[0043] Step S1013: Perform pulse compression processing based on the echo signal to obtain the pulse compression result. Optionally, the pulse compression result can be obtained by multiplying the spectrum of the echo signal with the spectrum of the reference signal and then performing a Fourier transform. The pulse compression result is the radar range imaging model under narrowband interference environment, expressed as follows: ,in, For reference signal, This is for convolution processing.

[0044] In one optional embodiment of this application, the noise narrowband interference signal includes a noise amplitude modulation interference signal, a noise frequency modulation interference signal, and a noise phase modulation interference signal;

[0045] The noise amplitude modulation interference signal is ,in, It is a constant. The center frequency of the interference signal. This is a frequency-limited noise signal;

[0046] The noise frequency modulation interference signal is ,in, It is a constant. For frequency modulation coefficients, This is a frequency-limited noise signal;

[0047] Noise phase modulation interference signal is ,in, The phase modulation coefficient, This is a frequency-limited noise signal.

[0048] Specifically, the echo signals received by the ISAR system include narrowband noise interference signals. Optionally, the narrowband noise interference signals include amplitude modulation noise interference signals, frequency modulation noise interference signals, and phase modulation noise interference signals. All three types of noise interference signals include frequency-limited noise signals, and all of them change with the frequency-limited noise signals.

[0049] In one optional embodiment of this application, the frequency-limited noise signal is , The bandwidth of the frequency-limited noise signal is... ,and , For the real part of the band-limited noise signal, The imaginary part of the band-limited noise signal. i The imaginary unit, t For time.

[0050] Specifically, in this embodiment, a frequency-limited noise signal is obtained. Optionally, the frequency-limited noise signal is obtained by Gaussian noise low-pass filtering, and the noise narrowband interference signal in the above embodiment changes with the frequency-limited noise signal.

[0051] In an optional embodiment of this application, step S102 involves obtaining the spectrum of the echo signal based on the radar range imaging model and constructing the amplitude modulation coefficient vector of the echo signal spectrum. Further, an anti-narrowband interference waveform optimization model is constructed. The specific process relating the anti-narrowband interference waveform optimization model to the amplitude modulation coefficient vector of the echo signal spectrum is as follows:

[0052] Step S1021: Based on the radar range imaging model, acquire the target range image of the echo signal, transform the compressed result of the target range image of the echo signal to the frequency domain, and discretize it. Its expression is: ,in, To observe the target echo signal spectrum vector, For the spectrum vector of the narrowband noise interference signal, To observe the spectral vector of the noise signal, The spectral vector of the reference signal, This refers to the discrete unit number at the distance. The distance is the number of sampling points; the compressed result of the target distance image is expressed as: ,in, This is the inverse Fourier transform;

[0053] Step S1022: Construct the amplitude modulation coefficient vector of the echo signal. ,and , m =0,1,... M -1;

[0054] Step S1023: Obtain the entropy value of the pulse compression result of the echo signal. Use the entropy value of the pulse compression result of the echo signal as the objective function to evaluate the quality of the amplitude modulation coefficient vector. The expression for the entropy value of the pulse compression result of the echo signal is: ,in, The scattering intensity density is the result of pulse compression of the echo signal. , For total energy, , This represents the number of variables to be optimized in the amplitude modulation coefficient vector;

[0055] Step S1024: Construct a narrowband interference-resistant waveform optimization model based on the minimum entropy criterion. The narrowband interference-resistant waveform optimization model is as follows: .

[0056] Specifically, in this embodiment, an anti-narrowband interference waveform optimization model is constructed based on the radar range imaging model. For the ISAR system, the transmitted echo signal is known, and the pulse compression... The echo signal is perfectly matched with the echo signal of the observed target, that is, the echo signal of the observed target is exactly the same as the echo signal of the observed target. After pulse compression, it becomes an ideal sinc function; however, it is a narrowband interference signal. Observation noise and Mismatch will increase the noise floor of the pulse compression results for the observed target, affecting the detection, structural analysis, and subsequent processing of the observed target. This is especially true for narrowband interference signals, which are inherently high-energy and, after pulse compression, may even overwhelm the echo signal of the observed target, making it difficult to obtain high-resolution range images and two-dimensional imaging results. Alternatively, if the amplitude-frequency characteristic of the linear frequency modulated signal is directly used as the amplitude-frequency characteristic of the transmitted echo signal, then during the matched filtering process, because the amplitude-frequency characteristic of the corresponding frequency band where interference exists in the echo signal is large, after pulse compression, the interference term... The interference still exists and cannot be ignored, severely affecting the imaging of the observed target. In this embodiment, the detection and suppression of narrowband interference can be transformed into the estimation problem of the "0-1" amplitude modulation coefficient vector of the echo signal, that is, constructing a frequency-domain notch receiver filter. The notch frequency band of the receiver filter coincides with the frequency band where the narrowband interference is located, thus allowing the interference term to be filtered out during the matched filtering stage. Filtering; the following is a detailed process description:

[0057] Based on the radar range imaging model, the target range image of the echo signal is acquired. The compressed result of the target range image of the echo signal is transformed to the frequency domain and then discretized. Its expression is as follows:

[0058] ,in, These are the spectral vectors of the echo signal, the observed target echo signal, the narrowband noise interference signal, the observed noise signal, and the matched signal, respectively. This refers to the discrete unit number at the distance. The distance is the number of sampling points; the compressed result of the target distance image is expressed as: ,in, As shown in the inverse Fourier transform, the discretized expression shows that the amplitude modulation coefficient vector has a value of 0 in the frequency band where the narrowband interference is located, and a value of 1 in other positions. This can effectively suppress narrowband signal interference and improve the quality of distance imaging results.

[0059] Step S1022: Construct the amplitude modulation coefficient vector of the echo signal. ,and , m =0,1,... M -1;

[0060] Step S1023: To effectively evaluate the suppression effect, the entropy value of the pulse compression result of the echo signal is obtained. The entropy value of the pulse compression result of the echo signal is used as the objective function to evaluate the merits of the amplitude modulation coefficient vector. The expression for the entropy value of the pulse compression result of the echo signal is: ,in, The scattering intensity density is the result of pulse compression of the echo signal. , For total energy, , Let be the number of variables to be optimized in the amplitude modulation coefficient vector; according to matched filtering theory, when the amplitude modulation coefficient vector... When the modulation coefficient of the frequency band where the interference is located is 0, the interference term in the matched filter result is zero, the interference signal is suppressed, the focusing degree of the observed target range image is improved, and the entropy value is minimized; thus, the entropy function... The magnitude of the amplitude modulation coefficient vector is directly related to the imaging quality of the observed target, i.e., the effectiveness of interference suppression. Therefore, the anti-interference waveform optimization problem can be transformed into an amplitude modulation coefficient vector problem. The optimization problem;

[0061] Step S1024: Construct a narrowband interference-resistant waveform optimization model based on the minimum entropy criterion. The narrowband interference-resistant waveform optimization model is as follows: .

[0062] By following the steps above, high-quality, high-resolution range images can be obtained.

[0063] Figure 2 This is a flowchart of the multi-narrowband interference detection algorithm based on the GA-PSO sequential optimization algorithm provided in this embodiment of the invention. Please refer to it. Figure 2 As shown, in an optional embodiment of this application, the spectrum of the echo signal is divided into several sub-bands of equal length, the amplitude modulation coefficient of each sub-band is set to the same value, and the amplitude modulation coefficient of each sub-band is solved using the GA algorithm. The specific process of coarsely estimating the sub-band where the narrowband interference is located is as follows:

[0064] Step S1031: Divide the spectrum of the echo signal into uniform... Joke frequency band, and The dimension of the amplitude modulation coefficient vector to be optimized is... Downgraded to Then the reduced amplitude modulation coefficient vector is ,in, , The variables to be optimized in the amplitude modulation coefficient vector are... Convert to ,and ;

[0065] Step S1032: Use the GA algorithm to... Perform optimization to obtain the optimal solution. The corresponding amplitude modulation coefficient vector can then be obtained. Selecting entropy As the fitness of each individual, the roulette wheel selection criterion is used for individual selection, and the entropy function is applied. Make modifications to make it suitable for use in roulette individual selection fitness models. ,in, The value is a positive constant to ensure non-negative fitness for the amplitude modulation coefficients after dimensionality reduction. Optimization is represented as ;

[0066] Step S1033: Use the GA algorithm to optimize and solve the problem to obtain the optimal amplitude modulation coefficient. This allows us to obtain a coarse estimate of the sub-frequency band where narrowband interference is located.

[0067] For details, please refer to [link / reference]. Figure 2 As shown, in this implementation, the number of variables to be optimized in the amplitude modulation coefficient vector of the sub-band is... The number of sampling points is consistent with the number of echo distance sampling points, resulting in high dimensionality. Directly applying an algorithm for solution would be difficult to converge. Therefore, the amplitude modulation coefficient vector optimization of the sub-band needs to be divided into two steps. One step involves dividing the spectrum of the echo signal into equally spaced... Each sub-band has an amplitude modulation coefficient of either 0 or 1. The number of sub-bands corresponding to 0 represents the number of narrowband interferences, further reducing the dimensionality of optimization variables. Optionally, the upper limit for estimating the number of narrowband interferences is... One, of which This indicates rounding down; to reduce the dimensionality of optimization variables, it is usually taken as... Furthermore, since the amplitude modulation coefficient can only take two values, 0 or 1, it satisfies the requirements of a small number of variables and discrete solutions in both the dimension of optimization variables and the dimension of feasible solutions. This perfectly matches the scenarios that the GA algorithm excels at handling. Therefore, the amplitude modulation coefficient vector of the sub-band is optimized and divided above, and the GA algorithm is used to optimize the amplitude modulation coefficient of the echo signal spectrum.

[0068] Please continue to refer to this. Figure 2 As shown, in an optional embodiment of this application, step S1033 uses the GA algorithm for optimization to obtain the optimal amplitude modulation coefficient. The specific process is as follows:

[0069] Step 1: Set the initial algebra Crossover probability Probability of mutation Randomly initialize the population Population size is The maximum number of iterations is 1000, and the iteration convergence threshold is... ,in, This represents the maximum fitness value in the population.

[0070] Step 2: Calculate the fitness value for each individual. Determine if the iteration convergence threshold has been reached; if so, the algorithm terminates and outputs the best-performing individual in the population; otherwise, proceed to the next step.

[0071] Step 3: Use roulette wheel to select individuals from the population. The probability of being selected is Among them, the best individual in the current population is selected. If the best individual is better than the best individual in the past, the best individual in the past is replaced by the best individual in the past. At the same time, the best individual in the past is replaced by the worst individual in the current population.

[0072] Step 4: Perform crossover operations on the population. Individuals are paired up randomly, with probability. The crossover operation is performed, and the ratio is determined by two uniformly distributed random numbers. A gene segment about to cross over will generate A new individual;

[0073] Step 5: For each individual in the population, use probability... Perform a mutation operation, using a uniformly distributed random number as a proportion to determine the mutation point location, and invert the values ​​at the individual mutation points; let the number of iterations... If so, proceed to step 2 above.

[0074] It should be noted that the above results are obtained by running the GA algorithm, and the parameters can be adjusted according to actual needs.

[0075] Please continue to refer to this. Figure 2 As shown, in an optional embodiment of this application, the PSO algorithm is used to calculate the proportion of each sub-band in the spectrum of the echo signal, thereby achieving a precise estimation of the frequency band where narrowband interference occurs. The specific process is as follows:

[0076] S1041. Using the PSO algorithm, the number of particles in the PSO algorithm is set to... The first in the PSO algorithm group The particle after passing the first After the next iteration, the velocity and position are respectively and The position of the particle can be converted into the final amplitude modulation coefficient vector using the following formula:

[0077] ;

[0078] The fitness value of the individual is calculated using the entropy function, and the estimate of the sub-frequency band where the narrowband interference is located is expressed as follows: , Represents an individual in the PSO algorithm;

[0079] Step S1042: Solve using the PSO algorithm to obtain... The optimal solution can be used to obtain a precise estimate of the frequency band where the narrowband interference is located.

[0080] For details, please refer to [link / reference]. Figure 2 As shown, in this embodiment, the number of variables to be optimized in the amplitude modulation coefficient vector of the sub-band is . The number of sampling points is consistent with the number of echo distance sampling points, resulting in high dimensionality. Directly applying the algorithm for solution would be difficult to converge. Therefore, the amplitude modulation coefficient vector optimization of the sub-band needs to be divided into two steps. One step is to optimize the width of each segment in the sub-band, continuously approximating the true frequency band of the narrowband interference. The number of narrowband interferences and their frequency bands within the bandwidth under equal-length sub-band division are already determined. However, under this equal-length division, some signal energy outside the narrowband interference will inevitably be suppressed by the notch filter. Although the narrowband interference is also suppressed, it will affect the final imaging quality. Therefore, it is necessary to accurately estimate the bandwidth of each sub-band. To obtain more accurate results, real numbers are used to describe the proportion of each sub-band in the spectrum. The PSO algorithm can handle continuous variable optimization problems well, so the PSO algorithm is used to further optimize the bandwidth of the above frequency bands.

[0081] Please continue to refer to this. Figure 2 As shown, in an optional embodiment of this application, the PSO algorithm is used in step S1042 to solve the problem, and the optimal solution obtained is used to obtain the precise estimation result of the frequency band where the narrowband interference is located. The specific process is as follows:

[0082] Step 1: Set the initial algebra Inertial weight Randomly initialize particle swarm positions and speed Population size is The maximum number of iterations is Save GA encoded output ;

[0083] Step 2: Output according to the GA algorithm. Convert the value of each particle into an amplitude modulation coefficient vector.

[0084] ;

[0085] Based on the amplitude modulation coefficient vector obtained by conversion Calculate the fitness value for each particle. ;

[0086] Step 3: The fitness value of each particle and the best position experienced by that particle. Compare the fitness values; if it is better than The fitness value is used as the best position for the current individual;

[0087] Step 4: The fitness value of each particle and the best position the particle has experienced globally. Compare the fitness values; if it is better than The fitness value is used as the current global best position;

[0088] Step 5: Using the formula and The velocity and position of the particles are updated, where, , The acceleration constant, and for A random number that is uniformly distributed on the upper surface;

[0089] Step 6, let ,like Then the process ends, and the amplitude modulation coefficient vector of the optimal individual is output. Otherwise If so, proceed to step 2 above.

[0090] It should be noted that when running the above results using the PSO algorithm, the parameters can be adjusted according to actual needs.

[0091] In one optional embodiment of this application, the specific process of constructing a spectrum notch filter based on the estimated structure of the frequency band where the narrowband interference is located, and then achieving interference suppression after matched filtering, is as follows:

[0092] By using the amplitude modulation coefficient vector as the amplitude modulation term of the matched filter while keeping the phase unchanged, the echo signal is matched and filtered to achieve interference suppression and waveform optimization.

[0093] Specifically, in this embodiment, the amplitude modulation coefficients of the sub-bands obtained by the GA algorithm and the PSO algorithm are used as the amplitude frequency characteristics of the receiving filter, and the phase frequency characteristics of the linear frequency modulated signal are used as the phase frequency characteristics of the receiving filter. By using this receiving filter to perform matched filtering on the echo signal, effective suppression of narrowband interference and waveform optimization can be achieved, thereby further realizing target focusing imaging.

[0094] In another embodiment of this application, the effectiveness of the method in the embodiment of this application is verified through simulation experiments:

[0095] 1. Simulation conditions

[0096] Figure 3 This is a schematic diagram of a two-dimensional point target model of an aircraft provided in an embodiment of the present invention. Please refer to it. Figure 3As shown in the simulation, the radar operates in the X-band with a bandwidth of 400MHz, a sampling frequency of 480MHz, a pulse repetition frequency of 100Hz, an azimuth imaging accumulation pulse count of 512 points, an imaging accumulation angle of 4°, and an interference-to-signal ratio of 25dB. The narrowband interference frequency bands are -152~-148MHz, -52~-47MHz, 69-73MHz, and 148~152MHz, respectively. The first two are noise phase modulation interference, and the latter two are noise frequency modulation interference and noise amplitude modulation interference, respectively.

[0097] 2. Simulation Experiment Content and Result Analysis

[0098] Figure 4(a) is a schematic diagram comparing the optimized waveform and the interference power spectrum provided in this embodiment of the invention. Referring to Figure 4(a), the optimized waveform power spectrum forms a depression at the interference point, with frequency bands of -152 to -147.5 MHz, -52.5 to -47 MHz, 68.5 to 73 MHz, and 147.5 to 152 MHz respectively. Figure 4(b) is a schematic diagram comparing the one-dimensional range image of isolated scattering points of the optimized waveform and the unoptimized waveform provided in this embodiment of the invention. Referring to Figure 4(b), the one-dimensional range image of the optimized waveform is not affected by interference, and the point target is very clear. Figure 4(c) is a two-dimensional imaging result of the unoptimized waveform and the optimized waveform provided in this embodiment of the invention, and Figure 4(d) is another two-dimensional imaging result of the unoptimized waveform and the optimized waveform provided in this embodiment of the invention. Referring to Figures 4(c) and 4(d), because the interference signal was not filtered out during pulse compression of the unoptimized waveform, it was severely affected, and the target was submerged in interference. Imaging with the optimized waveform of this invention can effectively suppress interference and obtain a focused imaging result. As can be seen from the above analysis, the algorithm proposed in this invention can effectively detect the number of interferences, carrier frequency and bandwidth, and can generate corresponding amplitude modulation coefficients. It can filter out interference in the matched filtering stage and still obtain a focused two-dimensional ISAR image in a multi-narrowband interference environment.

[0099] It should be noted that those skilled in the art will understand that all or part of the steps of the above method embodiments can be implemented by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When the program is executed, it performs the steps of the above method embodiments. The aforementioned storage medium includes various media that can store program code, such as ROM, RAM, magnetic disk, or optical disk.

[0100] The above description, in conjunction with specific preferred embodiments, provides a further detailed explanation of the present invention. It should not be construed that the specific implementation of the present invention is limited to these descriptions. For those skilled in the art, various simple deductions or substitutions can be made without departing from the concept of the present invention, and all such modifications and substitutions should be considered within the scope of protection of the present invention.

Claims

1. A sequentially optimized ISAR multi-narrowband interference detection and suppression method, characterized in that, include: Step S101: Acquire echo signals and construct a radar range imaging model under narrowband interference environment; Step S102: Based on the radar range imaging model, obtain the spectrum of the echo signal, and construct the amplitude modulation coefficient vector of the echo signal spectrum. Further, construct an anti-narrowband interference waveform optimization model; wherein, the anti-narrowband interference waveform optimization model is related to the amplitude modulation coefficient vector of the echo signal spectrum. Step S103: Divide the spectrum of the echo signal into several sub-bands of equal length, set the amplitude modulation coefficient of each sub-band to the same value, and use the GA algorithm to solve the amplitude modulation coefficient of each sub-band to make a coarse estimate of the sub-band where the narrowband interference is located; wherein, the amplitude modulation coefficient vector of the spectrum of the echo signal includes multiple amplitude modulation coefficients, and each sub-band corresponds to one amplitude modulation coefficient; Step S104: Use the PSO algorithm to solve for the proportion of each sub-frequency band in the spectrum of the echo signal, and make a precise estimate of the frequency band where the narrowband interference is located; wherein, the frequency band is the frequency range of the narrowband interference in the spectrum of the echo signal. Step S105: Construct a spectrum notch filter based on the estimation results of the frequency band where the narrowband interference is located, and achieve interference suppression after matched filtering.

2. The sequentially optimized ISAR multi-narrowband interference detection and suppression method according to claim 1, characterized in that, The specific process of acquiring the echo signal and constructing the radar range imaging model under narrowband interference environment in step S101 is as follows: Step S1011: Obtain the noise narrowband interference signal ; Step S1012: Receive the echo signal , The echo signal contains the noise narrowband interference signal, wherein, To observe the noise, To observe the echo signal of the target, L represents the number of equivalent scattering centers of the echo signal from the observed target. For the first Backscattering coefficient of each scattering center For distance window, The width of the radar-transmitted pulse signal is [value missing], and the bandwidth of the pulse signal is [value missing]. , The first echo signal of the observed target The instantaneous distance between the scattering center and the radar. The speed of electromagnetic wave propagation. The carrier frequency of the pulse signal transmitted by the radar. The frequency modulation frequency of the broadband signal transmitted by the radar; Step S1013: Perform pulse compression processing on the echo signal to obtain a radar range imaging model under narrowband interference environment; the radar range imaging model is... ,in, For reference signal, This is for convolution processing.

3. The sequentially optimized ISAR multi-narrowband interference detection and suppression method according to claim 2, characterized in that, The noise narrowband interference signal includes noise amplitude modulation interference signal, noise frequency modulation interference signal and noise phase modulation interference signal; The noise amplitude modulation interference signal is ,in, It is a constant. The center frequency of the interference signal. This is a frequency-limited noise signal; The noise frequency modulation interference signal is ,in, It is a constant. For frequency modulation coefficients, This is a frequency-limited noise signal; The noise phase modulation interference signal is ,in, The phase modulation coefficient, This is a frequency-limited noise signal.

4. The sequentially optimized ISAR multi-narrowband interference detection and suppression method according to claim 3, characterized in that, The frequency-limited noise signal is , Wherein, the bandwidth of the frequency-limited noise signal is ,and , The real part of the frequency-limited noise signal, The imaginary part of the band-limited noise signal is... i The imaginary unit, t For time.

5. The sequentially optimized ISAR multi-narrowband interference detection and suppression method according to claim 4, characterized in that, In step S102, based on the radar range imaging model, the spectrum of the echo signal is obtained, and the amplitude modulation coefficient vector of the echo signal spectrum is constructed. Further, an anti-narrowband interference waveform optimization model is constructed. The specific process relating the anti-narrowband interference waveform optimization model to the amplitude modulation coefficient vector of the echo signal spectrum is as follows: Step S1021: Based on the radar range imaging model, acquire the target range image of the echo signal, transform the compression result of the target range image of the echo signal to the frequency domain, and discretize it. Its expression is: ,in, The spectrum vector of the observed target echo signal, The frequency spectrum vector of the noise narrowband interference signal, To observe the spectral vector of the noise signal, The spectral vector of the reference signal, This refers to the discrete unit number at the distance. The distance sampling point is represented as: The compression result of the target distance image is expressed as: ,in, This is the inverse Fourier transform; Step S1022: Construct the amplitude modulation coefficient vector of the spectrum of the echo signal. ,and , m =0,1,... M -1; Step S1023: Obtain the pulse compression result entropy value of the echo signal, and use the pulse compression result entropy value of the echo signal as the objective function to evaluate the quality of the amplitude modulation coefficient vector, wherein the expression for the pulse compression result entropy value of the echo signal is: ,in, The scattering intensity density is the result of pulse compression of the echo signal. , For total energy, , The number of variables to be optimized in the amplitude modulation coefficient vector; Step S1024: Construct a narrowband interference-resistant waveform optimization model based on the minimum entropy criterion. The narrowband interference-resistant waveform optimization model is as follows: .

6. The sequentially optimized ISAR multi-narrowband interference detection and suppression method according to claim 5, characterized in that, The specific process of dividing the spectrum of the echo signal into several equal-length sub-bands, setting the amplitude modulation coefficient of each sub-band to the same value, and using the GA algorithm to solve for the amplitude modulation coefficient of each sub-band to coarsely estimate the sub-band where the narrowband interference is located is as follows: Step S1031: Divide the spectrum of the echo signal into uniform... The sub-frequency band mentioned in the segment, and The dimension of the amplitude modulation coefficient vector to be optimized is determined by... Downgraded to Then the reduced amplitude modulation coefficient vector is ,in, , The variables to be optimized in the amplitude modulation coefficient vector are... Convert to ,and ; Step S1032: Use the GA algorithm to... Perform optimization to obtain the optimal solution. The corresponding amplitude modulation coefficient vector can then be obtained. Selecting entropy As the fitness of each individual, the roulette wheel selection criterion is used for individual selection, and the entropy function is applied. Make modifications to make it suitable for use in roulette individual selection fitness models. ,in, For positive constants, the amplitude modulation coefficients after dimensionality reduction Optimization is represented as ; Step S1033: Use the GA algorithm to optimize and solve the problem to obtain the optimal amplitude modulation coefficient. This allows us to obtain a coarse estimate of the sub-frequency band where the narrowband interference is located.

7. The sequentially optimized ISAR multi-narrowband interference detection and suppression method according to claim 6, characterized in that, In step S1033, the GA algorithm is used for optimization to obtain the optimal amplitude modulation coefficient. The specific process is as follows: Step 1: Set the initial algebra Crossover probability Probability of mutation Randomly initialize the population Population size is The maximum number of iterations is 1000, and the iteration convergence threshold is... ,in, This represents the maximum fitness value in the population. Step 2: Calculate the fitness value for each individual. Determine if the iteration convergence threshold has been reached; if so, the algorithm terminates and outputs the best-performing individual in the population; otherwise, proceed to the next step. Step 3: Use roulette wheel to select individuals from the population. The probability of being selected is Among them, the best individual in the current population is selected. If the best individual is better than the best individual in the past, the best individual in the past is replaced by the best individual in the past. At the same time, the best individual in the past is replaced by the worst individual in the current population. Step 4: Perform crossover operations on the population. Individuals are paired up randomly, with probability. The crossover operation is performed, and the ratio is determined by two uniformly distributed random numbers. A gene segment about to cross over will generate A new individual; Step 5: For each individual in the population, use probability... Perform a mutation operation, using a uniformly distributed random number as a proportion to determine the mutation point location, and invert the values ​​at the individual mutation points; let the number of iterations... Then proceed to step 2.

8. The sequentially optimized ISAR multi-narrowband interference detection and suppression method according to claim 7, characterized in that, In step S104, the PSO algorithm is used to calculate the proportion of each sub-frequency band in the spectrum of the echo signal, and the specific process of accurately estimating the frequency band where the narrowband interference is located is as follows: S1041. Using the PSO algorithm, the number of particles in the PSO algorithm is set to... The first in the PSO algorithm group The particle after passing the first After the next iteration, the velocity and position are respectively and The position of the particle can be converted into the final amplitude modulation coefficient vector by the following formula: ； The fitness value of the individual is calculated using the entropy function, and the estimate of the sub-frequency band where the narrowband interference is located is expressed as follows: , Represents an individual in the PSO algorithm; Step S1042: Solve using the PSO algorithm to obtain... The optimal solution can be used to obtain the precise estimation result of the frequency band where the narrowband interference is located.

9. The sequentially optimized ISAR multi-narrowband interference detection and suppression method according to claim 8, characterized in that, The specific process of obtaining the optimal solution obtained by using the PSO algorithm in step S1042, which yields the precise estimation result of the frequency band where the narrowband interference is located, is as follows: Step 1: Set the initial algebra Inertial weight Randomly initialize particle swarm positions and speed Population size is The maximum number of iterations is Save GA encoded output ; Step 2: Output according to the GA algorithm. The value of each particle is converted into the amplitude modulation coefficient vector. ； Based on the amplitude modulation coefficient vector obtained by conversion Calculate the fitness value for each particle. ; Step 3: The fitness value of each particle and the best position experienced by that particle. Compare the fitness values; if it is better than The fitness value is used as the best position for the current individual; Step 4: The fitness value of each particle and the best position the particle has experienced globally. Compare the fitness values; if it is better than The fitness value is used as the current global best position; Step 5: Using the formula and The velocity and position of the particles are updated, where, , The acceleration constant, and for A random number that is uniformly distributed on the upper surface; Step 6, let ,like Then the process ends, and the amplitude modulation coefficient vector of the optimal individual is output. Otherwise Then proceed to step 2.

10. The sequentially optimized ISAR multi-narrowband interference detection and suppression method according to claim 1, characterized in that, The specific process of constructing a spectrum notch filter based on the estimation results of the frequency band where the narrowband interference is located, and then achieving interference suppression after matched filtering, is as follows: The amplitude modulation coefficient vector is used as the amplitude modulation term of the matched filter, and the phase remains unchanged. Matched filtering is then applied to the echo signal to achieve interference suppression and waveform optimization.

Images