Intermittent sampling forwarding jamming identification suppression method and device based on instance segmentation

Abstract

The application discloses a kind of intermittent sampling forwarding interference identification inhibition method and device based on instance segmentation, the method includes: the time-frequency chart of echo signal is input to the interference identification model based on mask trained to obtain the predicted mask information of each signal, wherein, predicted mask information includes mask position and mask value, and the mask value of different kinds of signals is different, and the label of sample interference signal and sample target signal in the sample mask time-frequency chart used for training interference identification model is made one by one;The predicted mask information of interference signal is mapped to time-frequency domain, and the time-frequency data of interference signal is obtained;According to the time-frequency data of interference signal, interference signal is inhibited.The application can realize the identification of interference signal one by one, and enhance interference identification inhibition effect.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of signal processing, and particularly relates to a method and device for identifying and suppressing intermittent sampling and forwarding jamming based on instance segmentation. BACKGROUND

[0002] With the development of digital radio frequency memory, coherent false target jamming technology means can be realized. The original intention of the ingenious intermittent sampling and forwarding jamming technology is to solve the problem of antenna spatial isolation in engineering application by using time-sharing transmission and reception. Due to the advantages of signal coherence, fast response, diverse jamming pattern and rich modulation pattern, the intermittent sampling and forwarding jamming is widely used in radar countermeasure technology research. In the research of radar anti-jamming technology, the radar jamming identification and suppression technology in the single intermittent sampling and forwarding jamming scene has made certain progress, but the research in the complex multi-intermittent sampling and forwarding jamming scene is still insufficient. In the increasingly complex jamming environment, the jamming identification and suppression technology in the multi-jamming scene needs to be developed.

[0003] At present, the jamming suppression technology for intermittent sampling and forwarding jamming is mainly based on the time domain discontinuous characteristics of the jamming. The radar received signal can be subjected to short-time Fourier transform, fractional order transform, short-time fractional order transform, etc., to obtain more obvious jamming and target signal characteristics. Referring to Figure 1 , the distribution of radar echo signals in time domain and frequency domain can be seen. In the time domain, the target and the jamming, and the jamming and the jamming are mixed with each other and have no obvious distinguishability. Referring to Figure 2 After converting the time domain data to the time-frequency domain, due to the time-sharing transmission and reception characteristics of the intermittent sampling and forwarding jamming, the jamming presents a discontinuous slice-like distribution in the time-frequency domain, which is obviously different from the continuous inclined target signal. Therefore, the jamming can be suppressed according to the characteristics of the jamming in the time-frequency domain.

[0004] The current jamming suppression method based on the time-frequency domain mainly identifies the jamming signal according to the number of peak values of the echo signal in the time-frequency domain, the 3dB width of the peak value and the number of distance gates where the peak value exists. This method can only identify all the jamming signals as a whole, and cannot specifically identify which jamming signal the time-frequency data belongs to. At the same time, the jamming identification accuracy is low when the signal-to-noise ratio is lower than-5dB.

[0005] Therefore, the current jamming identification and suppression method has poor jamming identification effect and suppression effect. SUMMARY

[0006] The embodiments of the present application provide a method and device for identifying and suppressing intermittent sampling and forwarding jamming based on instance segmentation, which can solve the problem of poor jamming identification effect and suppression effect of the current jamming identification and suppression method.

[0007] In a first aspect, an embodiment of the present application provides a method for identifying and suppressing intermittent sampling and forwarding interference based on instance segmentation, the method comprising:

[0008] inputting the time-frequency graph of the echo signal into the trained mask-based interference identification model to obtain predicted mask information of each signal, wherein the predicted mask information comprises a mask position and a mask value, the mask values of different types of signals are different, and the labels of the sample interference signals and the sample target signals in the sample mask time-frequency graph used for training the interference identification model are made one by one;

[0009] mapping the predicted mask information of the interference signal to the time-frequency domain to obtain time-frequency data of the interference signal;

[0010] suppressing the interference signal according to the time-frequency data of the interference signal.

[0011] In a second aspect, an embodiment of the present application provides an apparatus for identifying and suppressing intermittent sampling and forwarding interference based on instance segmentation, the apparatus comprising:

[0012] an interference identification module, the interference identification module comprising a trained mask-based interference identification model, the interference identification model being configured to input the time-frequency graph of the echo signal into the trained mask-based interference identification model to obtain predicted mask information of each signal, wherein the predicted mask information comprises a mask position and a mask value, the mask values of different types of signals are different, and the labels of the sample interference signals and the sample target signals in the sample mask time-frequency graph used for training the interference identification model are made one by one;

[0013] an interference information processing module, the interference information processing module being configured to map the predicted mask information of the interference signal to the time-frequency domain to obtain time-frequency data of the interference signal;

[0014] an interference suppression module, the interference suppression module being configured to suppress the interference signal according to the time-frequency data of the interference signal.

[0015] In a third aspect, an embodiment of the present application provides an electronic device, comprising a processor and a memory, wherein the memory is configured to store a computer program, and the processor is configured to execute the computer program (instructions) stored in the memory to implement the method of the first aspect.

[0016] In a fourth aspect, an embodiment of the present application provides a computer readable storage medium, the computer readable storage medium storing a computer program, and when the computer program is executed, the method of the first aspect can be implemented.

[0017] Compared with the prior art, the embodiment of the present application has the beneficial effects that: according to the method provided by the present application, the interference recognition model can extract more accurate features of the interference signal, and improve the recognition accuracy of the interference signal; by using samples in which the information of each interference signal is known in training the interference recognition model, the interference recognition model can output the prediction mask information of each interference signal for indicating the time-frequency information when used, so as to realize the identification of the interference signal one by one, and enhance the recognition effect; thereby enhancing the suppression effect of the interference signal. BRIEF DESCRIPTION OF DRAWINGS

[0018] Figure 1 A signal distribution diagram of a radar echo signal in a time domain is provided for the embodiment of the present application.

[0019] Figure 2 A signal distribution diagram of a radar echo signal in a time-frequency domain is provided for the embodiment of the present application.

[0020] Figure 3 An implementation flowchart of a training method of a mask-based interference recognition model is provided for the embodiment of the present application.

[0021] Figure 4 A sample time-frequency diagram and a sample mask time-frequency diagram are provided for the embodiment of the present application.

[0022] Figure 5 A training scene diagram of an interference recognition model is provided for the embodiment of the present application.

[0023] Figure 6 An implementation flowchart of an intermittent sampling and forwarding interference recognition and suppression method based on instance segmentation is provided for the embodiment of the present application.

[0024] Figure 7 A time-frequency diagram, signal classification result and prediction mask information of an echo signal are provided for the embodiment of the present application.

[0025] Figure 8 A structure diagram of an intermittent sampling and forwarding interference recognition and suppression device based on instance segmentation is provided for the embodiment of the present application.

[0026] Figure 9 A diagram of a pulse pressure result is provided for the embodiment of the present application.

[0027] Figure 10 Another time-frequency diagram, signal classification result and prediction mask information of an echo signal are provided for the embodiment of the present application.

[0028] Figure 11 Another diagram of a pulse pressure result is provided for the embodiment of the present application.

[0029] Figure 12 A diagram of interference suppression accuracy is provided for an embodiment of the present application.

[0030] Figure 13 A structural diagram of an electronic device is provided for an embodiment of the present application. DETAILED DESCRIPTION

[0031] In the following description, for purposes of explanation and not limitation, specific details are set forth such as particular architectures, techniques, etc. in order to provide a thorough understanding of embodiments of the present application. However, it will be apparent to those skilled in the art that the present application can be practiced in other embodiments that depart from these specific details. In other instances, detailed descriptions of well-known methods, devices, circuits, and

[0032] It will be understood that the terms "comprises" and / or "comprising," when used in this specification, specify the presence of stated features, integers, steps, operations, elements, and / or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and / or groups thereof.

[0033] It is also to be understood that the term "and / or" as used herein refers to and encompasses any and all possible combinations of one or more of the associated listed items, and that the term "at least one of' denotes one, or a plurality of, or any combination of the listed items.

[0034] As used in the description of the application and the appended claims, the term "if' can be interpreted to mean "when" or "upon" or "in response to determining" or "in response to detecting," depending on the context. Similarly, the phrase "if it is determined" or "if [a described condition or event] is detected" can be interpreted to mean "upon determining" or "in response to determining" or "upon [the described condition or event] being detected" or "in response to [the described condition or event] being detected," depending on the context.

[0035] In addition, the terms "first," "second," "third," etc. are used herein only to describe different instances of elements, and are not intended to imply or suggest relative importance of the elements.

[0036] Reference in the specification to "one embodiment" or "some embodiments" means that a particular feature, structure, or characteristic described in connection with the embodiment is included in at least one embodiment of the application. The appearances of the phrase "in one embodiment" or "in some embodiments" in various places in the specification are not necessarily all referring to the same embodiment, however, but can refer to one or more but less than all of the possible special features, structures, or characteristics of that embodiment. The terms "including," "containing," "having," and "encompassing" are used herein in the sense of "including, but not limited to," unless otherwise specifically noted.

[0037] The current interference identification and suppression method can perform interference identification and suppression through the following steps S201-S204.

[0038] S201, pulse compression processing and short-time Fourier transform processing are performed on the echo signal to obtain time-frequency domain data of the echo signal.

[0039] S202, based on the time-frequency domain data of the echo signal, the number M of branches within each range gate is analyzed p , the 3dB width ΔF of the peak value, and the number N of range gates in which the peak value exists p .

[0040] S203, interference signal identification is performed according to the above parameters.

[0041] Specifically, if the 3dB width of the peak value is approximately equal to the pulse width of the transmitted signal, it is determined as the target signal. If the 3dB width of the peak value is less than the pulse width of the transmitted signal and the number of peak values in each range gate is equal, it is judged according to the number N of range gates in which the peak value exists. If N p =1, it is intermittent sampling direct repeater jamming, if N p >1, it is intermittent sampling repeated repeater jamming. If the 3dB width of the peak value is less than the pulse width of the transmitted signal and the number of peak values in each range gate is not equal, it is intermittent sampling cyclic repeater jamming.

[0042] S204, the time-frequency domain time dimension of the interference signal and the target signal is determined, and interference is suppressed based on a time-frequency domain filter.

[0043] For example, the time-frequency domain filter can satisfy the following formula:

[0044] F=[α1α2…α N ]

[0045] F is the time-frequency domain filter, α i is the column vector of the time-frequency domain filter at the i-th time dimension, and N is the number of time-frequency domain time dimension sampling points.

[0046] wherein:

[0047]

[0048] I is a unit column vector, r is the distance gate where the target signal is located, p min is the minimum amplitude of the alpha i is the minimum amplitude of the alpha i is the i-th time dimension column vector.

[0049] As an example, the time-frequency domain filter is multiplied by the time-frequency distribution of the pulse pressure result to realize filtering. In this process, the signal in the time dimension where the target is located remains unchanged; and the signal in the time dimension where the interference is located is equal to the minimum value of the column, so as to realize interference suppression.

[0050] The interference identification method has low interference identification accuracy under the condition that the signal-to-noise ratio is lower than-5dB in a single intermittent sampling and forwarding interference scene, and it is also difficult to identify each interference in a multiple intermittent sampling and forwarding interference scene.

[0051] Therefore, the present application provides an intermittent sampling and forwarding interference identification and suppression method based on instance segmentation, which can extract more accurate features of interference signals through an interference identification model, improve the identification accuracy of interference signals, and output each interference signal for indicating the predicted mask information of time-frequency information through the interference identification model when in use, so as to realize the identification of interference signals one by one and enhance the identification effect, thereby enhancing the suppression effect of interference signals.

[0052] The present application will be further described in detail below in conjunction with specific embodiments, but the embodiments of the present application are not limited thereto.

[0053] The wireless blockchain network slicing method provided by the embodiments of the present application can be applied to electronic devices such as mobile terminals, personal notebook computers, supercomputers, etc., and the embodiments of the present application do not make any limitation on the specific type of electronic devices.

[0054] Figure 3 An implementation flowchart of a training method of a mask-based interference identification model provided by the embodiments of the present application is shown. As an example but not limitation, the training method can include steps S301-S306, which can be applied to the above-mentioned electronic devices, and each step will be described below.

[0055] S301, high-dimensional semantic information of a sample time-frequency diagram is extracted, and the high-dimensional semantic information of the sample time-frequency diagram and low-dimensional semantic information of the sample time-frequency diagram are fused to obtain semantic features of the sample time-frequency diagram.

[0056] Exemplarily, the sample time-frequency diagram can include time-frequency data of at least one sample interference signal.

[0057] For example, referring to the sample time-frequency diagram in (a) of FIG. 1, three intermittent sampling interference signals are included. Figure 4

[0058] Exemplarily, different labels are made for each signal in the sample echo signal, for example, the target is marked as 1, the noise is marked as 0, and the interference is marked as 2, and then the sample mask time-frequency diagram can be obtained by marking different signals with labels in the sample time-frequency diagram of the sample echo signal.

[0059] Generally, the number of noise signals and target signals is 1, and the number of interference signals can be a positive integer n. j Therefore, the sample time-frequency diagram includes n all = 2 + n j signals.

[0060] For example, referring to the sample mask time-frequency diagram in (b) of FIG. 1, a label mask is made according to the labels of the signals in the diagram, the label values are different, the color values of the label masks are different, and different color slices represent different signals. Among them, the noise is blue-purple and blended in the background, the three interference signals are light blue, dark yellow and light yellow respectively, and the target is dark blue. Figure 4 In a possible implementation, the sample time-frequency diagram can be drawn according to two-dimensional time-frequency domain data of the echo signal, and the two-dimensional time-frequency domain data is obtained by performing a short-time Fourier transform on the sample echo signal.

[0061] Exemplarily, the two-dimensional time-frequency domain data can satisfy the following formula:

[0062]

[0063]

[0064] Wherein, STFT(n, f) is the two-dimensional time-frequency domain data, t0 is the sliding window interval, n is the sliding window frequency, f is the frequency, S r (t) is the sample echo signal, w(t-nt0) is the window function, and j is the imaginary unit.

[0065] Wherein:

[0066] S r (t) = S e (t) + S j (t) + n0(t)

[0067]

[0068] S e (t) is the sample target signal in the sample echo signal, S​​j (t) is an interference part, n0(t) is noise in the sample echo signal, t1 is the initial start time of the sliding window, t2 is the initial end time of the sliding window, and Δt = t2-t1 is the window length of the window function.

[0069] In one example, the sample transmission signal for generating the sample echo signal can be a linear frequency modulation signal.

[0070] For example, the sample transmission signal can satisfy the following formula:

[0071]

[0072] where S(t) is the sample transmission signal, represents a rectangular window function, which represents the function value of the independent variable t in the range of and the function value is 0 in the remaining range; k = B / T P is the frequency modulation rate, B is the bandwidth of the sample transmission signal, and T p is the pulse width of the sample transmission signal.

[0073] In a possible implementation, referring to Figure 5 , the interference identification model can be a Mask Region-based Convolutional Neural Network (Mask R-CNN). It can include a backbone feature extraction network, a region proposal network (RPN), a proposal layer, a region of interest alignment (ROIAlign) layer, and three prediction branches.

[0074] In one example, referring to Figure 5 , the backbone feature extraction network can include a residual network (ResNet-50) and a feature pyramid network (FPN).

[0075] For example, the sample time-frequency diagram can be input into the interference identification model, the residual network can perform encoder dimension reduction on the m-1 level semantic information to extract high-dimensional semantic information to obtain the m level semantic information, and finally M-1 level semantic information is obtained. Then the feature pyramid can start from the 3rd level semantic information, up-sample the m level semantic information, and fuse with the m-1 level semantic information to obtain the m-1 level semantic feature, up-sample the M level semantic information to obtain the M level semantic feature, maximum pool the M level semantic feature to obtain the M+1 level semantic feature, and finally obtain M level semantic features.

[0076] For example, m is a positive integer greater than or equal to 2 and less than or equal to M. For example, M can be equal to 5.

[0077] For example, the second-level to the Mth-level semantic information can be feature maps with different dimensions, such as C2 to C5 in Figure 5 The first-level semantic information can be a sample time-frequency map.

[0078] For example, the high dimension and the low dimension of the semantic information are relative. In the (m-1)th-level semantic information and the mth-level semantic information, the (m-1)th-level semantic information is low-dimensional semantic information, and the mth-level semantic information is high-dimensional semantic information.

[0079] For example, the semantic feature can be a feature layer, see P2 to P6 in Figure 5

[0080] S302, predicting based on the semantic feature of the sample time-frequency map to obtain the change of the plurality of prediction boxes and the confidence of the signal contained in each prediction box.

[0081] For example, see Figure 5 The semantic feature of the sample time-frequency map can be input to the RPN. One branch passes through a convolution layer (for example, the orange filled part in Figure 5 and a normalized exponential function softmax to obtain the confidence of the signal contained in each prediction box on each pixel point on each feature layer. Another branch passes through a convolution layer and a bounding box regression algorithm (Bounding Boxes Regression, bbox reg) to obtain the change of each prediction box on each pixel point on the feature layer.

[0082] S303, screening and correcting the prediction box according to the confidence and the change of the prediction box to obtain the proposal box.

[0083] For example, the prediction box can be input to the proposal layer. The proposal layer can first select the prediction box with a confidence probability of the top 6000, then correct the position of the prediction box according to the change of the prediction box, discard the prediction box whose corrected bounding box exceeds the size of the sample time-frequency map. Finally, the top 2000 prediction boxes with the largest probability are obtained by the softmax function. The intersection over union of the 2000 prediction boxes and the corresponding real box information is calculated. If it is greater than 0.5, it is set as a positive sample, otherwise it is set as a negative sample. Then, 200 samples are selected as proposal boxes according to the ratio of positive and negative samples of 1:2.

[0084] S304, pooling the semantic feature of the sample time-frequency map and the sample time-frequency map in the proposal box to obtain a preset size semantic feature and a preset size local sample time-frequency map. ​

[0085] Exemplarily, the semantic features and the sample time-frequency diagram in the proposal box can be pooled by a pooling layer to obtain semantic features of a preset size and local sample time-frequency diagrams of a preset size.

[0086] S305, based on the semantic features of a preset size and the local sample time-frequency diagrams of a predicted size, the predicted mask information of each sample signal is obtained.

[0087] Exemplarily, the predicted mask information can include mask positions and mask values, and the mask values of different types of signals are different. The category of the signal can be indicated by the mask value.

[0088] Exemplarily, referring to Figure 5 The semantic features of a preset size and the local sample time-frequency diagrams of a predicted size can be respectively input into three prediction branches, the classification information of each sample signal is predicted by the full connection layer and the softmax function in the classification branch, the position of each signal, i.e. the position of the bounding box including the signal (i.e. the position of the mask), is predicted by the full connection layer and the bounding box regression algorithm in the regression branch, and the mask value of each signal is obtained by the deconvolution layer in the mask branch.

[0089] S306, the total loss is determined according to the predicted mask information of the sample echo signal and the real mask information in the sample mask time-frequency diagram, and the trained interference recognition model is obtained by training the interference recognition model according to the total loss.

[0090] Exemplarily, the total loss of the model can be obtained by comparing the predicted mask information of the sample echo signal and the real mask information in the sample mask time-frequency diagram, and then the trained interference recognition model is obtained by training the interference recognition model based on the gradient descent algorithm. Since the interference recognition model is based on instance segmentation to recognize the interference signal, and the total loss is calculated based on the sample mask time-frequency diagram which is respectively labeled for each interference signal, the recognition accuracy of the interference recognition model for the interference signal can be improved, and it is possible to identify the interference signal one by one.

[0091] Exemplarily, after obtaining the trained interference recognition model, the sample time-frequency diagram in the test set can also be input into the trained interference recognition model for recognition to obtain the predicted mask information, and whether the model is successfully trained is determined according to the predicted mask information of the sample time-frequency diagram in the test set and the real mask information in the sample mask time-frequency diagram. If they are the same, it is successful, and if they are different, it is a failure.

[0092] Optionally, the proportion of the sample time-frequency diagrams in the training set, the validation set and the test set can be 4:1:5.

[0093] Figure 6An implementation flowchart of an example segmentation-based intermittent sampling forwarding interference identification and suppression method provided by an embodiment of the present application is shown. The interference identification and suppression method can be applied in the electronic device described above, and can include steps S601-S603, which are described below.

[0094] S601, input the time-frequency graph of the echo signal into the trained mask-based interference identification model to obtain the predicted mask information of each signal.

[0095] For example, the interference identification model can be trained according to the training method described above. In use, the operation process of each module of the interference identification model is similar to that during training, except that the input is the time-frequency graph of the echo signal to be detected during use. The specific operation process of the interference identification model can be referred to the description of steps S301-S305 in the training method described above, and will not be described here.

[0096] For example, the interference identification model can obtain the classification result of each signal (see (b) in Figure 7 according to the time-frequency graph of the echo signal (see (a) in Figure 7 , and assign different mask values to different types of signals on the time-frequency graph according to the classification result to obtain the predicted mask information (see (c) in Figure 7 .

[0097] For example, referring to (b) in Figure 7 , the purple box represents the interference signal, and the red box represents the target signal. Different mask values are assigned to different signals according to their categories, and the mask values are represented by different colors to obtain (c) in Figure 7 . The light blue, dark yellow, and light yellow covered areas represent the areas of the three interference signals, and the dark blue lines represent the area of the target signal.

[0098] S602, map the predicted mask information of the interference signal to the time-frequency domain to obtain the time-frequency data of the interference signal.

[0099] For example, the mask position of the signal is the position of the mask on the time-frequency graph, and the time-frequency domain information of the signal can be obtained by mapping the mask position to the time-frequency domain.

[0100] S603, suppress the interference signal according to the time-frequency data of the interference signal.

[0101] In one example, a corresponding filter can be made according to the time-frequency data of the interference signal, and the interference can be filtered through the filter.

[0102] In one example, the positions of the mask of the interference signal on the time-frequency graph of the echo signal can be set to 0 to suppress the interference to obtain the time-frequency graph of the echo signal after interference suppression.

[0103] According to the interference identification and suppression method provided by the present application, the interference identification model can extract more accurate features of the interference signal, and improve the identification accuracy of the interference signal; by using samples in which the information of each interference signal is known in training the interference identification model, the interference identification model can output the prediction mask information of each interference signal for indicating the time-frequency information when used, so as to realize the identification of the interference signal one by one, and enhance the identification effect; thereby enhancing the suppression effect of the interference signal.

[0104] Figure 8 Fig. 1 shows a structural schematic diagram of an interference identification and suppression device based on instance segmentation provided by an embodiment of the present application. As an example but not limitation, the device 800 can include an interference identification module 810, an interference information processing module 820, and an interference suppression module 830.

[0105] The interference identification module 810 includes a trained mask-based interference identification model, which is used to input the time-frequency graph of the echo signal into the trained mask-based interference identification model to obtain the prediction mask information of each signal, wherein the prediction mask information includes a mask position and a mask value, and the mask values of different signals are different; the labels of the sample interference signals and the sample target signals in the sample mask time-frequency graph used for training the interference identification model are made one by one;

[0106] The interference information processing module 820 is used to map the prediction mask information of the interference signal to the time-frequency domain to obtain the time-frequency data of the interference signal.

[0107] The interference suppression module 830 is used to suppress the interference signal according to the time-frequency data of the interference signal.

[0108] In order to better illustrate the beneficial effects of the interference identification and suppression method provided by the present application, the following simulation experiment is performed.

[0109] Figure 9 Fig. 4 shows a schematic diagram of the pulse pressure result.

[0110] For example, the echo signal in this experiment is an echo signal with three intermittent sampling interference signals under the condition of 30 dB.

[0111] Referring to Figure 9 , Figure 9 Fig. 4(a) is the pulse pressure result of the echo signal, Figure 9 Fig. 4(b) is the pulse pressure result of the echo signal after interference suppression.

[0112] Based on the formula: JIR = p_jam-a_jam, where p_jam is the interference power of the echo signal, and a_jam is the interference power of the echo signal after interference suppression, the interference suppression ratio JIR is calculated according to the results of the pulse compression before and after interference suppression, and the JIR is obtained as 31.3363 dB.

[0113] Figure 11 Another schematic diagram of a pulse compression result provided by an embodiment of the present application is shown.

[0114] For example, the experiment is performed under the conditions of a signal-to-noise ratio of -10 dB and the presence of three interferences. The time-frequency diagram of the input echo signal is shown in (a) of FIG. 12, the pulse compression result of the echo signal is shown in (a) of FIG. 13, the classification result and the predicted mask information predicted by the interference identification model are shown in (b) and (c) of FIG. 14, respectively, and the pulse compression result of the echo signal after interference suppression is shown in (b) of FIG. 15. Figure 10 Figure 11 Figure 10 Figure 11

[0115] Through calculation, it can be obtained that the interference suppression ratio of the experiment reaches 27.1318%.

[0116] Figure 12 A schematic diagram of an interference suppression accuracy rate provided by an embodiment of the present application is shown.

[0117] Referring to FIG. 18, Figure 12 It can be seen that under the conditions that the signal-to-interference ratio is set to 20 dB, and the signal-to-noise ratio is set to -10 dB to 10 dB (with a step of 5 dB), the interference identification accuracy rate reaches more than 94%.

[0118] According to the interference identification and suppression method provided by the present application, the interference identification model can extract more accurate features of the interference signal, and improve the identification accuracy rate of the interference signal; by using the sample in which the information of each interference signal is known in training the interference identification model, the interference identification model can output the predicted mask information of each interference signal for indicating the time-frequency information when used, so as to realize the identification of the interference signals one by one, and enhance the identification effect; thereby enhancing the suppression effect on the interference signals.

[0119] Figure 13 A structural schematic diagram of an electronic device provided by an embodiment of the present application is shown. As shown in Figure 13 The electronic device 1300 shown in FIG. 13 can include at least one processor 1310 Figure 13 ​​​​The electronic device 1300 can be a processing device such as a robot, which can implement the above method. Embodiments of the present application do not limit the specific type of the electronic device.

[0120] The electronic device 1300 can be a processing device such as a robot, which can implement the above method. Embodiments of the present application do not limit the specific type of the electronic device.

[0121] Those skilled in the art can understand that Figure 13 The electronic device 1300 is only an example and does not limit the electronic device. The electronic device 1300 can include more or fewer components than shown, or can combine some components, or have different components. For example, the electronic device 1300 can also include an input / output interface.

[0122] The processor 1310 can be a central processing unit (CPU), and can also be other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic components, discrete hardware components, etc. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor.

[0123] The memory 1320 can be an internal storage unit, such as a hard disk or a memory, in some embodiments. The memory 1320 can also be an external storage device, such as a plug-in hard disk, a smart memory card (SMC), a secure digital (SD) card, a flash card, etc., in other embodiments. Further, the memory 1320 can include both an internal storage unit and an external storage device. The memory 1320 is used to store an operating system, application programs, a boot loader, data, and other programs, such as program codes of the computer program, etc. The memory 1320 can also be used to temporarily store data that has been output or will be output.

[0124] It should be understood that the magnitude of the serial number of each step in the above embodiments does not mean the order of execution, and the execution order of each process should be determined according to its function and inherent logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0125] It can be clearly understood by those skilled in the art that, for the convenience and brevity of description, only the above division of each functional unit and module is exemplified, and in actual application, the above functions can be completed by different functional units and modules according to needs, that is, the internal structure of the device is divided into different functional units or modules to complete all or part of the above described functions. Each functional unit and module in the embodiment can be integrated in one processing unit, or each unit can exist physically, or two or more units can be integrated in one unit. The integrated unit can be realized in the form of hardware or in the form of software functional unit. In addition, the specific name of each functional unit and module is only for convenient distinction, and does not limit the protection scope of the present application. The specific working process of the unit and module in the above system can refer to the corresponding process in the above method embodiments, and will not be repeated here.

[0126] The embodiment of the present application further provides a computer readable storage medium, the computer readable storage medium stores a computer program, and the computer program is executed by a processor to realize the steps in each method embodiment.

[0127] The embodiment of the present application provides a computer program product, when the computer program product runs on an electronic device, so that the electronic device executes to realize the steps in each method embodiment.

[0128] The integrated unit, if implemented in the form of a software function unit and sold or used as an independent product, can be stored in a computer readable storage medium. Based on such understanding, the present application can implement all or part of the processes in the above-mentioned embodiment methods, which can be completed by instructing relevant hardware through a computer program. The computer program can be stored in a computer readable storage medium, and the computer program can implement the steps of each method embodiment when executed by a processor. The computer program includes computer program code, which can be in the form of source code, object code, executable files or some intermediate forms. The computer readable medium can at least include any entity or device capable of carrying the computer program code to the photographing device / terminal equipment, recording medium, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signal, telecommunication signal and software distribution medium. For example, U disk, mobile hard disk, magnetic disk or optical disk, etc. In some jurisdictions, according to legislation and patent practice, the computer readable medium can not be an electrical carrier signal and a telecommunication signal.

[0129] In the above embodiments, the description of each embodiment has its own focus, and the parts not described or recorded in detail in a certain embodiment can be referred to the relevant description of other embodiments.

[0130] Those skilled in the art can realize that the units and algorithm steps of each example described in combination with the embodiments disclosed herein can be realized by electronic hardware or a combination of computer software and electronic hardware. Whether the functions are realized in hardware or software depends on the specific application and design constraints of the technical solution. Professional technicians can use different methods to realize the described functions for each specific application, but such implementation should not be considered beyond the scope of the present application.

Claims

1. An instance segmentation based intermittent sampling and forwarding jammer identification and mitigation method, characterized in that, The method comprises the following steps: inputting a time-frequency graph of an echo signal into a trained mask-based interference identification model to obtain predicted mask information of each signal, wherein the predicted mask information comprises a mask position and a mask value, and the mask values of different signals are different, and the labels of sample interference signals and sample target signals in a sample mask time-frequency graph used for training the interference identification model are made one by one; mapping the predicted mask information of the interference signal to the time-frequency domain to obtain time-frequency data of the interference signal; suppressing the interference signal according to the time-frequency data of the interference signal; wherein the interference identification model is used for: extracting high-dimensional semantic information of the time-frequency graph and fusing the high-dimensional semantic information with low-dimensional semantic information of the time-frequency graph to obtain semantic features of the time-frequency graph; predicting the change of a plurality of prediction boxes and the confidence of the signals contained in each prediction box based on the semantic features of the time-frequency graph; screening and correcting the prediction boxes according to the confidence and the change of the prediction boxes to obtain a suggestion box; pooling the semantic features of the time-frequency graph and the time-frequency graph in the suggestion box to obtain a preset size of semantic features and a preset size of local time-frequency graph; obtaining the predicted mask information of each signal based on the preset size of semantic features and the preset size of local time-frequency graph; determining the total loss according to the predicted mask information and the real mask information of the sample echo signal, and training the interference identification model according to the total loss to obtain the trained interference identification model; wherein the extraction of the high-dimensional semantic information of the time-frequency graph and the fusion of the high-dimensional semantic information with the low-dimensional semantic information of the time-frequency graph to obtain the semantic features of the time-frequency graph comprises: encoder dimension reduction is performed on the m-1 level semantic information to obtain the m level semantic information, wherein m is a positive integer greater than or equal to 2 and less than or equal to M, and in the m level semantic information and the m-1 level semantic information, the m-1 level semantic information is low-dimensional semantic information, and the m level semantic information is high-dimensional semantic information; starting from the third level semantic information, the m level semantic information is up-sampled and fused with the m-1 level semantic information to obtain the m-1 level semantic feature; the M level semantic information is up-sampled to obtain the M level semantic feature; the M level semantic feature is pooled to obtain the M+1 level semantic feature.

2. The method of claim 1, wherein, The sample mask time-frequency graph is obtained by adding a label mask of each signal to a sample time-frequency graph, and the sample time-frequency graph is drawn according to two-dimensional time-frequency domain data of a sample echo signal, and the two-dimensional time-frequency domain data is obtained by performing short-time Fourier transform on the sample echo signal.

3. The method of claim 2, wherein, The two-dimensional time-frequency domain data satisfies the following formula: wherein is the two-dimensional time-frequency domain data, is a sliding window interval, is a number of sliding times of the sliding window, is a frequency, is the sample echo signal, is a window function, j is an imaginary unit.

4. The method of claim 2, wherein, The sample transmission signal of the sample echo signal is a linear frequency modulation signal.

5. The method of claim 1, wherein, The suppression of the interference signal according to the time-frequency data of the interference signal comprises: zeroing the time-frequency data of the interference signal in the time-frequency data of the echo signal to suppress the interference signal.

6. An instance segmentation based intermittent sampling and forwarding jammer identification and suppression apparatus, characterized by, The method comprises the following steps: The interference identification module comprises a trained mask-based interference identification model, and the interference identification model is configured to input a time-frequency map of an echo signal into the trained mask-based interference identification model to obtain predicted mask information of each signal, wherein the predicted mask information comprises a mask position and a mask value, and the mask values of different signals are different, and labels of sample interference signals and sample target signals in a sample mask time-frequency map used for training the interference identification model are made one by one. The interference information processing module is configured to map the predicted mask information of the interference signal to a time-frequency domain to obtain time-frequency data of the interference signal. The interference suppression module is configured to suppress the interference signal according to the time-frequency data of the interference signal. The interference identification model is configured to: extract high-dimensional semantic information of the time-frequency map and fuse the high-dimensional semantic information and low-dimensional semantic information of the time-frequency map to obtain semantic features of the time-frequency map; predict a change of a plurality of prediction boxes and a confidence of a signal contained in each prediction box based on the semantic features of the time-frequency map; screen and correct the prediction boxes based on the confidence and the change of the prediction boxes to obtain a suggestion box; pool the semantic features of the time-frequency map and a time-frequency map in the suggestion box to obtain a preset size of semantic features and a preset size of local time-frequency maps; obtain the predicted mask information of each signal based on the preset size of semantic features and the preset size of local time-frequency maps; determine a total loss based on the predicted mask information and real mask information of sample echo signals, and train the interference identification model based on the total loss to obtain the trained interference identification model; The extraction of the high-dimensional semantic information of the time-frequency map and the fusion of the high-dimensional semantic information and the low-dimensional semantic information of the time-frequency map to obtain the semantic features of the time-frequency map comprises: encoder dimension reduction of m-1 level semantic information to obtain m level semantic information, wherein m is a positive integer greater than or equal to 2 and less than or equal to M, and in the m-1 level semantic information and the m level semantic information, the m-1 level semantic information is low-dimensional semantic information, and the m level semantic information is high-dimensional semantic information; starting from the third level semantic information, upsampling of the m level semantic information and fusion with the m-1 level semantic information to obtain m-1 level semantic features; upsampling of the M level semantic information to obtain M level semantic features; pooling of the M level semantic features to obtain M+1 level semantic features.

7. An electronic device comprising a memory, a processor, and a computer program stored in the memory, characterized in that The processor executes the computer program to implement the method of any one of claims 1-5.

8. A computer-readable storage medium storing a computer program, the computer-readable storage medium comprising: The computer program is executed by the electronic device to implement the method of any one of claims 1-5.

Images