CN110187313B

CN110187313B - Radar signal sorting and identifying method and device based on fractional order Fourier transform

Info

Publication number: CN110187313B
Application number: CN201910470154.XA
Authority: CN
Inventors: 王功明; 陈世文; 黄洁; 齐艳丽; 邢小鹏; 秦鑫; 胡雪若白; 苑军见
Original assignee: Information Engineering University of PLA Strategic Support Force
Current assignee: Information Engineering University of PLA Strategic Support Force
Priority date: 2019-05-31
Filing date: 2019-05-31
Publication date: 2021-05-07
Anticipated expiration: 2039-05-31
Also published as: CN110187313A

Abstract

The invention belongs to the technical field of radar signal identification, and particularly relates to a radar signal sorting and identifying method and device based on fractional order Fourier transform, wherein the method comprises the following steps: acquiring a fractional order domain oscillogram of each radar signal by using fractional order Fourier transform; respectively carrying out normalization processing on the oscillograms, and constructing a combined characteristic parameter set; and utilizing a support vector machine classifier to perform sorting identification on the feature vectors in the combined feature parameter set. According to the method, the difference of different signals in fractional order domain waveforms is utilized, and a plurality of groups of symmetrical Holder coefficient values are extracted to form a combined characteristic vector, so that automatic sorting and identification of radar signals are realized; the method has better similar aggregation degree and stronger anti-noise performance, can completely express the time-frequency characteristic of a signal fractional order domain, and has certain application prospect; simulation experiment results show that when the signal-to-noise ratio is greater than 4dB, the overall average identification success rate of 6 radar signals is better than 96.5%, and when the signal-to-noise ratio is 0dB, the overall average identification success rate of 6 radar signals is better than 84.8%; the method also has a good identification effect on the composite modulation signal.

Description

Radar signal sorting and identifying method and device based on fractional order Fourier transform

Technical Field

The invention belongs to the technical field of radar signal identification, and particularly relates to a radar signal sorting and identifying method and device based on fractional Fourier transform.

Background

Radar signal sorting identification is an important component of electronic reconnaissance and support. With the development of modern radar technology, the modulation modes of signals are diversified, so that the traditional sorting and identifying method based on Pulse Description Word (PDW) is difficult to adapt to the requirement of signal identification, and therefore, the analysis and mining of features in radar signals are gradually a research hotspot. In recent years, scholars at home and abroad do a lot of work on extraction of features in radar signals, and various analysis methods are proposed, such as a time-frequency analysis method, a fuzzy function method, a high-order statistics method and the like. However, these methods are easily affected by noise under the condition of low signal-to-noise ratio, are difficult to effectively extract features, have limited application range, and can only process specific signals. Fractional Fourier Transform (FRFT) is used as a novel generalized time-frequency analysis method, can be regarded as the representation of a signal on a Fractional domain formed by rotating a coordinate axis on a time-frequency surface around an origin in an anticlockwise direction by any angle, not only retains the excellent characteristics of the Fourier Transform, but also has the unique advantages of the FRFT, and therefore, a better processing effect can be obtained under certain conditions. Modulation characteristic identification of linear frequency modulation signals, frequency coding signals and phase coding signals is completed based on the characteristic that the FRFT modulus changes along with the order and the waveform of the FRFT when the order is 1, but the signal types identified by the method are limited; by calculating the phase image coefficients and the vergence characteristics of FRFTs of different orders and further extracting the order value and the kurtosis corresponding to the peak value of the curve, a group of combined characteristic vectors are constructed for sorting and identifying 3 radar signals, and the method is sensitive to noise and has poor identification performance under low signal-to-noise ratio; the method for modulating and identifying the low-interception radar signal based on the FRFT is further provided by researching the energy peak distribution characteristics of the linear frequency modulation signal, the symmetrical triangular frequency modulation continuous wave signal and the multiphase coding signal in the fractional order domain, the flow of the method is complex, and the method cannot effectively process the radar signal with the complex modulation type.

Disclosure of Invention

Therefore, the invention provides a radar signal sorting and identifying method and device based on fractional order Fourier transform, which utilize the difference of different signals in fractional order domain waveforms to obtain the symmetrical Holder coefficient values to form a combined characteristic vector, realize automatic sorting and identifying of radar signals, and have better sorting and identifying performance and strong application prospect.

According to the design scheme provided by the invention, the radar signal sorting and identifying method based on fractional Fourier transform comprises the following contents:

A) acquiring a fractional order domain oscillogram of each radar signal by using fractional order Fourier transform;

B) respectively carrying out normalization processing on the oscillograms, and constructing a combined characteristic parameter set;

C) and utilizing a support vector machine classifier to perform sorting identification on the feature vectors in the combined feature parameter set.

In the step a), the continuous signal in the radar signal is divided into a convolution signal based on chirp, and fractional fourier transform is performed to obtain a radar signal fractional domain oscillogram.

The above, B) includes the following:

B1) normalizing the oscillogram to obtain a discrete sequence;

B2) introducing a basic signal to construct a basic signal sequence;

B3) selecting orders of a plurality of fractional Fourier transform, and respectively calculating corresponding symmetrical Holder coefficient values to form a combined feature vector;

B4) and obtaining a combined feature vector parameter set according to the intra-class aggregation degree and the inter-class dispersion degree.

Preferably, in B2), the basic signal is a rectangular signal and/or a triangular signal.

Preferably, B2), the rectangular signal sequence is represented as:

the triangular signal sequence is represented as:

wherein N is the length of the discrete sequence.

In the above, C), the method further includes a training and testing phase of the support vector machine classifier, where the training and testing phase includes the following contents:

C1) acquiring signal sample data, and dividing the signal sample data into a training sample for training and a test sample for testing;

C2) training and learning the support vector machine classifier by using the training samples;

C3) and testing by using the trained support vector machine classifier aiming at the test sample so as to carry out radar signal sorting and identification.

Preferably, the signal sample data comprises a plurality of types of radar signals under different signal-to-noise ratios, and the complex modulation signals are supplemented in the training samples for training and learning.

Furthermore, the invention also provides a radar signal sorting and identifying device based on fractional order Fourier transform, which comprises:

the acquisition module is used for acquiring a fractional order domain oscillogram of each radar signal by using fractional order Fourier transform;

the processing module is used for respectively carrying out normalization processing on the oscillogram and constructing a combined characteristic parameter set;

and the identification module is used for sorting and identifying the characteristic vectors in the combined characteristic parameter set by utilizing the support vector machine classifier.

In the above apparatus, the processing module includes:

the preprocessing submodule is used for carrying out normalization processing on the oscillogram to obtain a discrete sequence;

the construction submodule is used for introducing a basic signal and constructing a basic signal sequence;

the calculation submodule is used for selecting a plurality of orders of fractional Fourier transform, and respectively calculating corresponding symmetrical Holder coefficient values to form a combined feature vector;

and the obtaining submodule is used for obtaining the combined feature vector parameter set according to the intra-class aggregation degree and the inter-class dispersion degree.

In the above apparatus, the identification module comprises:

the training submodule is used for acquiring signal sample data and dividing the signal sample data into a training sample for training and a test sample for testing; training and learning the support vector machine classifier by using the training samples; testing the test sample by using the trained support vector machine classifier so as to carry out radar signal sorting and identification;

and the recognition submodule is used for sorting and recognizing the characteristic vectors in the combined characteristic parameter set by using the trained and tested support vector machine classifier.

The invention has the beneficial effects that:

in the invention, aiming at extracting intra-pulse characteristic parameters capable of reflecting intrinsic information of radar signals as effective supplement of five conventional parameters, the situation that the radar signals of the current new system are difficult to sort and identify is solved, and the like, a plurality of groups of symmetrical Holder coefficient values are extracted to form a combined characteristic vector by utilizing the difference of different signals in fractional order domain waveforms, so that the automatic sorting and identification of the radar signals are realized; the method has good similar aggregation degree and strong anti-noise performance, can completely express the time-frequency characteristic of the signal fractional order domain, and has certain application prospect. Further, simulation experiment results show that when the signal-to-noise ratio is greater than 4dB, the overall average identification success rate of 6 radar signals is superior to 96.5%, and when the signal-to-noise ratio is 0dB, the overall average identification success rate of 6 radar signals is superior to 84.8%; the method also has a good identification effect on the composite modulation signal.

Description of the drawings:

FIG. 1 is a flow chart of a radar signal sorting and identification method in an embodiment;

FIG. 2 is a diagram illustrating the construction of a joint feature parameter set according to an embodiment;

FIG. 3 is a flowchart of an embodiment of classifier training test;

FIG. 4 is a schematic diagram of a radar signal sorting and identifying device in an embodiment;

FIG. 5 is a schematic diagram of an exemplary processing module;

FIG. 6 is a schematic diagram of an identification module in an embodiment;

FIG. 7 is a schematic representation of the FRFT three-dimensional analysis of the LFM and BPSK signals at an SNR of 10dB in the example;

FIG. 8 is a FRFT waveform diagram of 8 radar signals of different orders under the condition of 10dB signal-to-noise ratio in the embodiment;

FIG. 9 is a schematic diagram of a flow chart of an implementation of a radar signal sorting identification algorithm in the embodiment;

FIG. 10 is a schematic diagram of recognition rates of training under different parameter values in the embodiment;

FIG. 11 is a diagram illustrating the identification accuracy of 6 radar signals with different SNR in the example;

FIG. 12 is a diagram showing the identification accuracy of 8 radar signals with different SNR in the example;

FIG. 13 is a comparison of average recognition accuracy of different algorithms in the embodiment.

The specific implementation mode is as follows:

in order to make the objects, technical solutions and advantages of the present invention clearer and more obvious, the present invention is further described in detail below with reference to the accompanying drawings and technical solutions.

In view of the situations of limited identification types, non-ideal performance, complex flow and the like in the existing radar signal identification, in the embodiment of the present invention, referring to fig. 1, a radar signal sorting and identifying method based on fractional order Fourier transform is provided, which includes the following contents:

s101) acquiring a fractional order domain oscillogram of each radar signal by using fractional order Fourier transform;

s102) respectively carrying out normalization processing on the oscillograms, and constructing a combined characteristic parameter set;

s103) utilizing a support vector machine classifier to perform sorting identification on the feature vectors in the combined feature parameter set.

Fractional order Fourier transformation is used for obtaining the fractional order domain waveform of the radar signal, a combined characteristic parameter set is constructed, automatic sorting and identification of the radar signal are achieved, and the problem that the radar signal of the current new system is difficult to sort and identify is solved.

Furthermore, in the embodiment of the invention, continuous signals in the radar signals are divided into convolution signals based on linear frequency modulation, fractional order Fourier transform is carried out to obtain the radar signal fractional order domain oscillogram, and the fractional order Fourier transform can be calculated by adopting fast Fourier transform with low calculation complexity, so that the operation performance is greatly improved.

Further, in the embodiment of the present invention, referring to fig. 2, the constructing of the joint feature parameter set includes the following contents:

s201) normalizing the oscillogram to obtain a discrete sequence;

s202) introducing a basic signal to construct a basic signal sequence;

s203) selecting orders of a plurality of fractional Fourier transforms, and respectively calculating corresponding symmetrical Holder coefficient values to form a combined eigenvector;

s204), obtaining a combined feature vector parameter set according to the intra-class aggregation degree and the inter-class dispersion degree.

And extracting a plurality of groups of symmetrical Holder coefficient values to form a combined characteristic vector by using the waveform difference of different signals in the fractional order domain, thereby realizing effective sorting and identification of various radar signal types.

Further, in the embodiment of the present invention, the basic signal is a rectangular signal and/or a triangular signal. Preferably, the rectangular signal sequence is represented as:

the triangular signal sequence is represented as:

wherein N is the length of the discrete sequence.

Further, in the embodiment of the present invention, as shown in fig. 3, a training and testing phase of a support vector machine classifier is further included, where the training and testing phase includes the following contents:

s301) acquiring signal sample data, and dividing the signal sample data into training samples for training and test samples for testing;

s302) training and learning the support vector machine classifier by using the training sample;

s303) testing the test sample by using the trained support vector machine classifier to perform radar signal sorting and identification.

Furthermore, an embodiment of the present invention further provides a radar signal sorting and identifying apparatus based on fractional Fourier transform, as shown in fig. 4, including:

an obtaining module 101, configured to obtain a fractional domain oscillogram of each radar signal by using fractional fourier transform;

the processing module 102 is configured to perform normalization processing on the oscillograms respectively and construct a joint feature parameter set;

and the identification module 103 is used for sorting and identifying the feature vectors in the combined feature parameter set by using a support vector machine classifier.

Further, in the above-mentioned embodiment of the apparatus, referring to fig. 5, the processing module 102 includes:

the preprocessing submodule 201 is used for performing normalization processing on the oscillogram to obtain a discrete sequence;

a construction submodule 202, configured to introduce a basic signal and construct a basic signal sequence;

the calculating submodule 203 is used for selecting a plurality of orders of fractional Fourier transform, and respectively calculating corresponding symmetrical Holder coefficient values to form a combined feature vector;

the obtaining sub-module 204 is configured to obtain a joint feature vector parameter set according to the intra-class aggregation degree and the inter-class dispersion degree.

Further, in the above embodiment of the apparatus, referring to fig. 6, the identification module 103 includes:

the training submodule 301 is configured to obtain signal sample data, and divide the signal sample data into a training sample for training and a test sample for testing; training and learning the support vector machine classifier by using the training samples; testing the test sample by using the trained support vector machine classifier so as to carry out radar signal sorting and identification;

and the identifying submodule 302 is configured to perform sorting and identifying on the feature vectors in the combined feature parameter set by using the trained support vector machine classifier.

In order to verify the effectiveness of the technical scheme in the embodiment of the invention, the following further explanation is made by combining various radar signals and simulation experiment data:

fractional Fourier transform can be understood as rotation of a time-frequency plane, can reflect information of signals in a time domain and a frequency domain at the same time, has no interference of cross terms compared with a common quadratic time-frequency analysis method, and is more suitable for processing non-stationary signals.

Let the signal be s (t), the most common FRFT is defined as follows:

wherein p is the order of FRFT, K_p(t, u) is the kernel function of FRFT, and its specific expression is

Where α ═ p pi/2 is the rotation angle of the FRFT,

is a complex exponential and n is an integer.

By variable substitution

And

substituting the kernel function into the FRFT definition equation

It can be seen that the period of the fractional Fourier transform is 4, and when p is 4n +1, i.e. (α is 2n pi + pi/2), the fractional Fourier transform is a conventional Fourier transform.

With the intensive research on the fractional Fourier transform fast algorithm, a plurality of fast discrete algorithms are proposed in recent years, so that the operational efficiency is greatly improved, the embodiment of the invention can adopt a decomposition type fast algorithm proposed by Ozaktas, the algorithm divides a continuous form into a convolution form based on a Chirp signal, and the FFT with low computational complexity can be adopted to calculate the FRFT through the expression, so that the discrete sampling method has the computational complexity almost the same as the FFT. Fig. 7 shows the results of FRFT three-dimensional analysis of the LFM signal and BPSK signal at a SNR of 10 dB. Fig. 8 shows FRFT waveforms of 6 typical radar signals and 2 complex modulation signals at different orders when the signal-to-noise ratio is 10dB, where the 8 signals are: conventional (CW) signals, Linear Frequency Modulation (LFM) signals, Nonlinear Frequency Modulation (NLFM) signals, Binary Phase Shift Keying (BPSK) signals, Quaternary Phase Shift Keying (QPSK) signals, Costas signals, LFM/FSK signals and BPSK/FSK composite Modulation signals, which correspond to (a), (b), (c), (d), (e), (f), (g) and (h) in fig. 8, respectively. As can be seen from fig. 8, the FRFT image shapes and distributions of different signals have obvious differences when p is 0.75,1.25, and 1.75, so that effective features can be extracted from the FRFT image for sorting and identifying radar radiation sources.

Let two one-dimensional discrete positive-valued signal sequences { S₁(i) ≧ 0, i ═ 1,2, …, N } and { S ≧ S₂(j) ≧ 0, j ≧ 1,2, …, N, if a, b > 1 and

the symmetric Holder coefficients of the two signals are

As can be seen from the definition of the symmetric Holder coefficient, when a equals 2, the symmetric Holder coefficient becomes the imaging coefficient.

In calculating the Holder coefficient, the FRFT waveform sequence is normalized to obtain { g (i) ═ 1,2, …, N }, where N is the length of the normalized waveform sequence. Two basic signal sequences are introduced, which can be respectively expressed as

Rectangular signal sequence:

triangular signal sequence:

from equations (5) and (6), the calculation equation for the symmetric Holder coefficient can be further derived as follows:

taking the order p of fractional Fourier transform as 0.75,1.25 and 1.75, respectively calculating the values of symmetrical Holder coefficients by formulas (7) and (8) to obtain a group of 6-dimensional joint feature vectors, namely [ H [ -H ]_u1,H_t1,H_u2,H_t2,H_u3,H_t3]. As can be seen from equations (7) and (8), if the values of a and b are selected differently, the symmetric Holder coefficients have different magnitudes. Therefore, the values of a and b must be properly adjusted, so that the extracted symmetric Holder coefficient characteristic has both good intra-class aggregation degree and large inter-class dispersion degree.

The implementation flow of the radar radiation source combined feature sorting identification algorithm constructed in the embodiment of the invention can be shown in fig. 9, and comprises the following steps: 1) performing FRFT conversion on a radar radiation source signal to obtain an FRFT oscillogram of the signal; 2) carrying out normalization processing on the FRFT oscillogram to obtain a discrete sequence G (i); 3) constructing a rectangular sequence U (k) and a triangular sequence T (k), taking the order p of fractional Fourier transform as 0.75,1.25 and 1.75, respectively calculating 3 sets of symmetrical Holder coefficient values to form a joint feature vector; 4) determining the values of the symmetrical Holder coefficients a and b; 5) and sorting and identifying the feature vectors by using a Support Vector Machine (SVM) classifier.

In a simulation experiment, 8 radar signals are sorted and identified, and in addition to CW, LFM, NLFM, BPSK, QPSK and Costas signals, 2 composite modulation signals LFM/FSK and BPSK/FSK are added. Because different radar signals have different parameters, for the convenience of description, the sampling frequency f is adopted_sUniformly distributed U (-) for example, U (1/8,1/4) indicates that the parameter ranges in [ f [ ]_s/8,f_s/4]A random number in between. Detailed test environment and parameter settings are shown in tables 1 and 2, and sampling frequency f is uniformly obtained_s64MHz, and a pulse width T of 16 mus.

TABLE 1 test Environment

Table 2 simulation parameter settings

And (3) making the signal-to-noise ratio SNR 10dB, and taking an integer of 2-10 for the parameter. The 8 signals in each combination produced 50 pulses each, and the results of the recognition rate obtained by training tests using the Classification receiver of the MATLAB itself are shown in FIG. 10. As can be seen from fig. 10, the training effect is best when a is 4, so that a is 4 and b is 4/3.

In a simulation experiment, the relationship between the identification accuracy and the signal-to-noise ratio is explored, and at first, 6 signals of CW, LFM, NLFM, BPSK, QPSK, Costas codes and the like are selected for testing. As the signal is inevitably interfered by noise in the transmission process, the noise is assumed to be white Gaussian noise, the range of the signal to noise ratio is-6 dB to 20dB, and the step is 2 dB. At each signal-to-noise ratio, 300 samples were generated for each signal, 200 for training and 100 for testing. The experimental result is shown in fig. 11, and it can be seen that, according to the technical scheme in the embodiment of the present invention, when the signal-to-noise ratio is greater than 8dB, the identification accuracy rates of 6 radar signals are all greater than 90%, and when the signal-to-noise ratio is greater than 10dB, the identification accuracy rates of other radar signals except QPSK all reach 100%. The technical scheme in the embodiment of the invention has better identification capability on LFM and Costas codes, and can obtain higher identification accuracy under the condition of lower signal-to-noise ratio. BPSK and QPSK signal identification accuracy decreases faster when the signal-to-noise ratio is below 2 dB. Along with the continuous reduction of the signal-to-noise ratio, the performance of the algorithm of the invention is also seriously reduced, which is mainly because the FRFT waveform is interfered by noise, the signal characteristics are not easy to extract, and the classification of the signal is difficult to complete well.

Training samples are further supplemented, 2 LFM/FSK and BPSK/FSK composite modulation signals which are widely applied to the radar of the new system are added, the identification accuracy of each signal under the condition of different signal-to-noise ratios is tested again, the experimental result is shown in figure 12, it can be seen that when the signal-to-noise ratio is 6dB, the identification success rate of 8 radar signals reaches 90%, and when the signal-to-noise ratio is 12dB or higher, the identification accuracy of other radar signals except QPSK reaches 100%. Experiments prove that the method has a good identification effect on the composite modulation radar signals.

The comparison results of the average identification rate experiments of 6 kinds of radar signals are shown in fig. 13, and document 4 refers to the existing radar radiation source signal sorting based on fuzzy function phase-like coefficients, the extracted phase-like coefficient characteristics are easily affected by noise, and the sorting identification results are sharply reduced when the signal-to-noise ratio is lower than 2 dB; document 5 indicates that the bispectrum features extracted by the existing complex system radar radiation source signal identification are insensitive to noise, but are affected by the box dimension and information dimension feature robustness extracted when the signal parameters change, and the sorting identification rate is relatively low; when the signal-to-noise ratio is 4dB, the overall average recognition rate of the technical scheme in the embodiment of the invention reaches 96.5%, and the noise-resistant performance is better under the condition of lower signal-to-noise ratio. Therefore, the characteristics provided by the invention based on the automatic sorting and identification of the fractional Fourier transform (FRFT) radar signals have better sorting and identification performance, the sorting and identification of 8 radar signals can be effectively realized, 3 groups of symmetrical Holder coefficient characteristics are extracted on the basis of the fractional domain waveforms of different orders, and the radar signals are automatically sorted and identified by constructing a combined feature vector and sending the combined feature vector to an SVM classifier. Experimental results show that the extracted feature vector has good similar aggregation degree and strong anti-noise performance, can completely express the time-frequency characteristic of a signal fractional order domain, and has a certain application prospect.

Unless specifically stated otherwise, the relative steps, numerical expressions, and values of the components and steps set forth in these embodiments do not limit the scope of the present invention.

The device provided by the embodiment of the present invention has the same implementation principle and technical effect as the method embodiments, and for the sake of brief description, reference may be made to the corresponding contents in the method embodiments without reference to the device embodiments.

It is clear to those skilled in the art that, for convenience and brevity of description, the specific working processes of the system and the apparatus described above may refer to the corresponding processes in the foregoing method embodiments, and are not described herein again.

In all examples shown and described herein, any particular value should be construed as merely exemplary, and not as a limitation, and thus other examples of example embodiments may have different values.

It should be noted that: like reference numbers and letters refer to like items in the following figures, and thus, once an item is defined in one figure, it need not be further defined and explained in subsequent figures.

The flowchart and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of systems, methods and computer program products according to various embodiments of the present invention. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of code, which comprises one or more executable instructions for implementing the specified logical function(s). It should also be noted that, in some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams and/or flowchart illustration, and combinations of blocks in the block diagrams and/or flowchart illustration, can be implemented by special purpose hardware-based systems which perform the specified functions or acts, or combinations of special purpose hardware and computer instructions.

Finally, it should be noted that: the above-mentioned embodiments are only specific embodiments of the present invention, which are used for illustrating the technical solutions of the present invention and not for limiting the same, and the protection scope of the present invention is not limited thereto, although the present invention is described in detail with reference to the foregoing embodiments, those skilled in the art should understand that: any person skilled in the art can modify or easily conceive the technical solutions described in the foregoing embodiments or equivalent substitutes for some technical features within the technical scope of the present disclosure; such modifications, changes or substitutions do not depart from the spirit and scope of the embodiments of the present invention, and they should be construed as being included therein. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.

Claims

1. A radar signal sorting and identifying method based on fractional Fourier transform is characterized by comprising the following steps:

C) sorting and identifying the characteristic vectors in the combined characteristic parameter set by using a support vector machine classifier;

A) dividing continuous signals in the radar signals into convolution signals based on linear frequency modulation, and performing fractional Fourier transform to obtain a radar signal fractional domain oscillogram;

B) comprises the following contents:

B1) normalizing the oscillogram to obtain a discrete sequence;

B2) introducing a basic signal to construct a basic signal sequence;

2. The radar signal sorting and identifying method based on the fractional Fourier transform of claim 1, wherein in B2), a rectangular signal and/or a triangular signal is adopted as a basic signal.

3. The radar signal sorting and identifying method based on fractional Fourier transform of claim 2, wherein in B2), the rectangular signal sequence is represented as:

the triangular signal sequence is represented as:

wherein N is the length of the discrete sequence.

4. The method for sorting and identifying radar signals based on fractional Fourier transform as claimed in claim 1, wherein C) further comprises a training and testing stage of a support vector machine classifier, wherein the training and testing stage comprises the following contents:

5. The method for radar signal sorting and identifying based on fractional Fourier transform of claim 4, wherein in C1), the signal sample data contains radar signals of multiple types with different signal-to-noise ratios, and a complex modulation signal is added to the training sample for training and learning.

6. A radar signal sorting and identifying device based on fractional Fourier transform, which is realized based on the method of claim 1 and comprises:

7. The apparatus for radar signal classification recognition based on fractional order Fourier transform of claim 6, wherein the processing module comprises:

8. The apparatus for radar signal classification identification based on fractional order Fourier transform of claim 6, wherein the identification module comprises: