WO2020215911A1 - 混合信号的分类方法、装置及电子设备 - Google Patents
混合信号的分类方法、装置及电子设备 Download PDFInfo
- Publication number
- WO2020215911A1 WO2020215911A1 PCT/CN2020/078573 CN2020078573W WO2020215911A1 WO 2020215911 A1 WO2020215911 A1 WO 2020215911A1 CN 2020078573 W CN2020078573 W CN 2020078573W WO 2020215911 A1 WO2020215911 A1 WO 2020215911A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- signal
- matrix
- mixed signal
- identified
- classified
- Prior art date
Links
- 238000000034 method Methods 0.000 title claims abstract description 80
- 239000011159 matrix material Substances 0.000 claims abstract description 163
- 238000000926 separation method Methods 0.000 claims abstract description 63
- 238000013145 classification model Methods 0.000 claims abstract description 35
- 238000012847 principal component analysis method Methods 0.000 claims abstract description 18
- 238000012706 support-vector machine Methods 0.000 claims description 31
- 238000004891 communication Methods 0.000 claims description 19
- 238000012549 training Methods 0.000 claims description 18
- 238000004364 calculation method Methods 0.000 claims description 14
- 238000012545 processing Methods 0.000 claims description 14
- 238000004590 computer program Methods 0.000 claims description 13
- 230000006870 function Effects 0.000 claims description 10
- 239000013598 vector Substances 0.000 claims description 6
- 238000010606 normalization Methods 0.000 claims description 5
- 238000004458 analytical method Methods 0.000 claims description 3
- 238000000354 decomposition reaction Methods 0.000 claims description 3
- 230000000875 corresponding effect Effects 0.000 description 66
- 238000000513 principal component analysis Methods 0.000 description 20
- 238000010586 diagram Methods 0.000 description 14
- 230000008569 process Effects 0.000 description 14
- 230000008054 signal transmission Effects 0.000 description 7
- 230000008859 change Effects 0.000 description 2
- 239000000203 mixture Substances 0.000 description 2
- 238000012360 testing method Methods 0.000 description 2
- 230000005540 biological transmission Effects 0.000 description 1
- 238000012512 characterization method Methods 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 230000002596 correlated effect Effects 0.000 description 1
- 238000013500 data storage Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 238000003062 neural network model Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 239000013307 optical fiber Substances 0.000 description 1
- 230000002093 peripheral effect Effects 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 239000004065 semiconductor Substances 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
- 238000007619 statistical method Methods 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
- G06V30/10—Character recognition
- G06V30/19—Recognition using electronic means
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/213—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods
- G06F18/2135—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods based on approximation criteria, e.g. principal component analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2411—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on the proximity to a decision surface, e.g. support vector machines
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/77—Processing image or video features in feature spaces; using data integration or data reduction, e.g. principal component analysis [PCA] or independent component analysis [ICA] or self-organising maps [SOM]; Blind source separation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/12—Classification; Matching
Definitions
- This application relates to the technical field of mixed signal classification, and in particular to a method, device and electronic device for classifying mixed signals.
- the signal receiving end After receiving the target signal mixed with the interference signal, the signal receiving end needs to classify the signals in the mixed signal to obtain the target signal. Since each signal corresponds to a specific modulation method, the various signals in the mixed signal are usually classified by identifying the modulation methods of the various signals in the mixed signal.
- the existing mixed signal classification methods usually use the following methods to classify the various signals in the mixed signal: the power of the various signals in the mixed signal is the same, or the signal strength of the target signal is much stronger than the other signals in the mixed signal At this time, the various signals in the mixed signal are separated by calculating the channel matrix of each signal in the mixed signal, and then the various signals in the mixed signal are classified by identifying the modulation mode of the various signals.
- the existing mixed signal classification method has strict requirements on the environment, which is embodied in that it is usually necessary to meet certain conditions before mixing Signals are classified, and the specific conditions include: the power of the various signals in the mixed signal is the same, or the signal strength of the target signal is much stronger than other signals in the mixed signal. It can be seen that the existing mixed signal classification method does not have universal applicability in practical applications, that is, the applicable scenarios are limited.
- the purpose of the embodiments of the present application is to provide a mixed signal classification method, device, and electronic equipment, so as to provide a mixed signal classification method with universal applicability.
- the specific technical solutions are as follows:
- an embodiment of the present application provides a method for classifying mixed signals.
- the method includes:
- the preset principal component analysis method PCA uses the preset principal component analysis method PCA to calculate the matrix corresponding to the mixed signal to obtain the mixed signal to be classified and determine the number of signal types contained in the mixed signal to be classified; the mixed signal to be classified is removed The mixed signal obtained after the noise in the mixed signal;
- the classification model is used to calculate and output the modulation mode of the signal to be identified according to the characteristics of the signal to be identified;
- the output result of the classification model includes: the modulation mode of the signal to be identified.
- the steps of calculating the matrix corresponding to the mixed signal by using a preset principal component analysis method PCA to obtain the mixed signal to be classified and determining the number of signal types contained in the mixed signal to be classified include:
- the mixed signal to be classified is obtained by calculation; the matrix corresponding to the mixed signal to be classified is
- the number of singular values that are not 0 among the singular values ⁇ 1 ,..., ⁇ N the number of signal types included in the mixed signal to be classified is determined.
- the step of determining the separation matrix according to the number of signal types contained in the mixed signal to be classified includes:
- the separation matrix w is obtained by calculation; wherein, the separation matrix w has M rows and N columns, M is the number of signal types contained in the mixed signal to be classified, N is the number of antennas that receive the mixed signal, and L is the signal Number of receptions, Represents the result of the mth column of the separation matrix w after k iterations, vector Is the mixed signal to be classified The i-th column.
- the step of using the separation matrix to separate each signal in the mixed signal to be classified to obtain the signal to be identified includes:
- the separation matrix The matrix corresponding to the mixed signal to be classified Multiply and separate each signal in the mixed signal to be classified to obtain the signal to be identified.
- the step of using the calculated high-order cumulant as the characteristic of the signal to be identified corresponding to the high-order cumulant includes:
- the normalized high-order cumulant is used as the characteristic of the signal to be identified corresponding to the high-order cumulant.
- the classification model is a support vector machine model
- the method further includes:
- the parameters in the current support vector machine are adjusted to obtain the support vector machine model.
- an embodiment of the present application provides a mixed signal classification device, which includes:
- a receiving module configured to receive a mixed signal, the mixed signal contains noise and at least two different signals
- the analysis module is used to calculate the matrix corresponding to the mixed signal by using the preset principal component analysis method PCA to obtain the mixed signal to be classified and determine the number of signal types contained in the mixed signal to be classified;
- the classified mixed signal is a mixed signal obtained after removing noise in the mixed signal;
- a determining module configured to determine a separation matrix according to the number of signal types contained in the mixed signal to be classified
- a separation module configured to use the separation matrix to separate each signal in the mixed signal to be classified to obtain the signal to be identified
- the calculation module is used to calculate the preset number of high-order cumulants corresponding to each signal to be identified in the signal to be identified;
- the feature module is configured to use the calculated high-order cumulant as the feature of the signal to be identified corresponding to the high-order cumulant;
- An input module configured to input the characteristics of the signal to be identified into a preset classification model; the classification model is used to calculate and output the modulation mode of the signal to be identified according to the characteristics of the signal to be identified;
- the obtaining module is used to obtain the output result of the classification model; the output result includes: the modulation mode of the signal to be identified.
- an embodiment of the present application provides an electronic device, including: a processor, a communication interface, a memory, and a communication bus, where the processor, the communication interface, and the memory complete mutual communication through the communication bus;
- Memory used to store computer programs
- the preset principal component analysis method PCA uses the preset principal component analysis method PCA to calculate the matrix corresponding to the mixed signal to obtain the mixed signal to be classified and determine the number of signal types contained in the mixed signal to be classified; the mixed signal to be classified is removed The mixed signal obtained after the noise in the mixed signal;
- the classification model is used to calculate and output the modulation mode of the signal to be identified according to the characteristics of the signal to be identified;
- the output result of the classification model includes: the modulation mode of the signal to be identified.
- an embodiment of the present application provides a computer-readable storage medium in which a computer program is stored, and when the computer program is executed by a processor, the following steps are implemented:
- the preset principal component analysis method PCA uses the preset principal component analysis method PCA to calculate the matrix corresponding to the mixed signal to obtain the mixed signal to be classified and determine the number of signal types contained in the mixed signal to be classified; the mixed signal to be classified is removed The mixed signal obtained after the noise in the mixed signal;
- the classification model is used to calculate and output the modulation mode of the signal to be identified according to the characteristics of the signal to be identified; obtaining the output result of the classification model;
- the output result includes: the modulation mode of the signal to be identified.
- the method, device and electronic equipment for classifying mixed signals provided by the embodiments of the present application can realize that after receiving the mixed signal, use the preset principal component analysis method PCA to calculate the matrix corresponding to the mixed signal to obtain The mixed signal to be classified and the number of signal types contained in the mixed signal to be classified is determined; the separation matrix is determined according to the number of signal types contained in the mixed signal to be classified; the separation matrix is used to separate the mixed signal to be classified Each signal in the signal to obtain the signal to be identified; respectively calculate the preset number of high-order cumulants corresponding to each signal to be identified in the signal to be identified; use the calculated high-order cumulants as the The characteristics of the signal to be identified corresponding to the high-order cumulant; input the characteristics of the signal to be identified into a preset classification model; and obtain the output result of the classification model.
- the mixed signal classification method provided in the embodiment of the application does not require the classification environment. Unlike the existing mixed signal classification method, it needs to meet specific conditions before the mixed signal can be classified; therefore, the embodiment of the application provides Compared with the prior art, the classification method of mixed signals has universal applicability.
- FIG. 1a is a schematic flowchart of a method for classifying mixed signals according to an embodiment of the application
- Figure 1b is a schematic diagram of a model of signal transmission and reception
- FIG. 2 is a schematic diagram of another process of a mixed signal classification method provided by an embodiment of the application.
- Figure 3 is a constellation diagram of a mixed signal formed by a mixture of signals QPSK and 16QAM received at the receiving end;
- 4 is a constellation diagram corresponding to the signal QPSK in the mixed signal obtained by processing the mixed signal without using the PCA provided in the embodiment of the present application;
- FIG. 5 is a constellation diagram corresponding to the signal QPSK in the mixed signal obtained after processing the mixed signal by using the PCA provided in an embodiment of the present application;
- FIG. 6 is a schematic structural diagram of a mixed signal classification device provided by an embodiment of the application.
- FIG. 7 is a schematic structural diagram of an electronic device provided by an embodiment of the application.
- the embodiments of the present application provide a mixed signal classification method, device and electronic device.
- the mixed signal classification method provided in the embodiment of the present application can be applied to the signal receiving end, and of course it is not limited to this.
- the mixed signal targeted by the present application is signal data containing multiple different signals obtained by mixing some interference signals with one signal. It is understandable that the interference signal is different from the noise in the signal transmission process, and the interference signal can change the transmitted signal from one type to multiple types.
- the embodiment of the present application provides a method for classifying mixed signals.
- the method includes:
- S101 Receive a mixed signal, and the mixed signal contains noise and at least 2 different signals;
- one signal source corresponds to one signal
- M signal sources can be used to send signals L times at the same time, and N antennas can receive these signals.
- M ⁇ N In order to ensure that the mixed signal can be completely received, M ⁇ N, where L can be a positive integer not less than 1.
- the received mixed signal is a signal obtained by mixing other signals in one signal transmission process, that is, a signal obtained by mixing signals from other signal sources during the signal transmission process of one signal source. .
- the so-called complex channel is a channel that transmits complex signals.
- the complex signal is a signal characterization method. Specifically, the modulus of the complex number is used to represent the amplitude of the signal, and the amplitude of the complex number is used to represent the phase of the signal.
- S102 Use the preset principal component analysis method PCA to calculate the matrix corresponding to the mixed signal to obtain the mixed signal to be classified and determine the number of signal types contained in the mixed signal to be classified; Mixed signal obtained after noise;
- PCA Principal Component Analysis
- PCA can be used to filter noise, because the influence of any component change is far greater than the influence of random noise, so for noise, various components are relatively unaffected, that is, Use principal components to reconstruct the original data.
- PCA can be used to calculate the matrix corresponding to the mixed signal, obtain the mixed signal to be classified and determine the specific implementation of the number of signal types contained in the mixed signal to be classified can be the same as the prior art. The application is not detailed here.
- the step of determining the separation matrix according to the number of signal types contained in the mixed signal to be classified may include:
- the separation matrix w has M rows and N columns, M is the number of signal types contained in the mixed signal to be classified, N is the number of antennas that receive the mixed signal, and L is the number of signal receptions, Represents the result of the m-th column of the separation matrix w after k iterations, vector Mixed signal to be classified The i-th column. It can be understood that L may also be referred to as the number of signal sampling times, and the value of L is the same as the number of signal transmissions.
- S104 Use a separation matrix to separate each signal in the mixed signal to be classified to obtain the signal to be identified;
- using the separation matrix to separate each signal in the mixed signal to be classified to obtain the signal to be identified is to use the separation matrix to separate various signals in the mixed signal to be classified to obtain a variety of signals to be identified.
- the step of using a separation matrix to separate each signal in the mixed signal to be classified to obtain the signal to be identified may include:
- the matrix The matrix corresponding to the mixed signal to be classified Multiply and separate each signal in the mixed signal to be classified to obtain the signal to be identified.
- S105 Calculate the preset number of high-order cumulants corresponding to each signal to be identified in the signals to be identified respectively;
- each signal to be identified has a preset number of high-order cumulants, and the process of calculating the preset number of high-order cumulants corresponding to each signal to be identified can refer to any of the prior art. A way to calculate the high-order cumulant of the signal is not repeated here.
- each kind of signal to be identified corresponds to a high-order cumulant. Then, for each signal to be identified, the high-order cumulant corresponding to the signal to be identified can be used as the corresponding signal to be identified. feature.
- the step of using the calculated high-order cumulant as the characteristic of the signal to be identified corresponding to the high-order cumulant may include:
- the normalized high-order cumulant is used as the characteristic of the signal to be identified corresponding to the high-order cumulant.
- the high-order cumulant corresponding to the signal to be identified is normalized, and the normalized high-order cumulant is used as the characteristic of the signal to be identified .
- the corresponding high-order cumulant of the signal to be identified may be determined as the characteristic of the signal to be identified.
- S107 Input the characteristics of the signal to be identified into a preset classification model; the classification model is used to calculate and output the modulation mode of the signal to be identified according to the characteristics of the signal to be identified;
- the classification model can be of multiple types, such as neural network model, or support vector machine model, and so on.
- S108 Obtain the output result of the classification model; the output result includes: the modulation mode of the signal to be identified.
- the classification of various signals in the mixed signal can be completed.
- the mixed signal classification method provided by the embodiment of the present application does not require the classification environment. Unlike the existing mixed signal classification method, it needs to meet specific conditions to perform the mixed signal classification. Classification; therefore, the mixed signal classification method provided by the embodiments of the present application has universal applicability compared with the prior art.
- the classification model may be a support vector machine model
- the method further includes:
- the parameters of the current support vector machine are adjusted to obtain the support vector machine model.
- the training sample is signal data, and the modulation method of the label signal data of the training sample.
- the support vector machine model can be Where the parameters Represents weight, parameter Represents the bias, It is the characteristic of the signal.
- the feature vectors corresponding to the training sample are The labels corresponding to the training samples are ⁇ 1 , ⁇ 2 ,..., ⁇ s , and the feature vector And its corresponding labels ⁇ 1 , ⁇ 2 ,..., ⁇ s , input the current support vector machine model to obtain the modulation mode corresponding to each training sample output by the current support vector machine model; according to the current support vector machine model output result and The label of the training sample, using the preset loss function:
- t s ⁇ 1,...,I ⁇ represents the training sample Belongs to the t sth modulation method, that is, the training sample It belongs to the t s- th signal; I represents how many modulation methods the support vector machine model can recognize, for example, I is 3, which means that the support vector machine model can recognize 3 modulation methods; i ⁇ 1,... ,I ⁇ t s ⁇ represents slack variable
- the coefficient of the penalty term is a constant used to control the degree of punishment for the wrong sample; Indicates the error value between the output result and the actual result.
- the coefficients of the slack variable and the penalty term may be set to values of 0.1 and 0.2, respectively.
- the formula Calculated according to the modulation method the characteristic corresponding to the modulation method.
- m test represents the modulation method; Represents the feature corresponding to m test ; the function argmax() is used to calculate the value of the parameter t corresponding to the maximum result value, which is to obtain the maximum possible modulation mode and the corresponding feature.
- the first step is to use the preset principal component analysis method PCA to remove the noise in the mixed signal to avoid the influence of noise on the classification result, and obtain the matrix corresponding to the mixed signal to be classified
- the second step is to determine the separation matrix w according to the number of signal types contained in the mixed signal to be classified, and use the separation matrix w to correspond to the matrix corresponding to the mixed signal to be classified obtained in the first step Multiply, reduce the correlation between various types of signals in the mixed signal to be classified, and obtain the signal to be identified.
- in order to reduce the amount of calculation usually separate the matrix w and the mixed signal to be classified Before multiplying, normalize the separation matrix w to get the matrix Reuse matrix The matrix corresponding to the mixed signal to be classified Multiply.
- the third step for each signal to be identified, calculate multiple high-order cumulants corresponding to the signal to be identified, use the calculated high-order cumulants as the characteristics of the signal to be identified, and input the characteristics of the signal to be identified into the preset Classification model: Support vector machine model, through the support vector machine model to identify the modulation format of the signal to be identified, so as to achieve the purpose of classifying mixed signals.
- S201 Use N antennas to receive the mixed signal; where N is a positive integer greater than 1;
- the number of antennas is usually greater than or equal to the number of signals that may be contained in the mixed signal.
- S205 Arrange the singular values ⁇ 1 ,..., ⁇ N from small to large; set the singular values of singular values ⁇ 1 ,..., ⁇ N whose values are less than the preset threshold value to 0, and calculate the diagonal matrix
- S207 Determine the number of signal types included in the mixed signal to be classified according to the number of singular values that are not 0 among the singular values ⁇ 1 ,..., ⁇ N ;
- the number of singular values that are not 0 is 3, so it is determined that the mixed signal to be classified contains 3 different signals.
- the number of iterations and the initial separation matrix can be set manually, for example, the number of iterations is set to 10 times, and the initial separation matrix is set to a unit matrix with M rows and N columns, where M is the received mixed signal containing The number of signal types, N is the number of antennas. For example, if the number of signal types included in the received mixed signal is 3 and the number of antennas is 10, the initial separation matrix is a unit matrix with 3 rows and 10 columns.
- the separation matrix w has M rows and N columns, M is the number of signal types contained in the mixed signal to be classified, N is the number of antennas that receive the mixed signal, and L is the number of signal receptions, Represents the result of the m-th column of the separation matrix w after k iterations, vector Mixed signal to be classified In the ith column of, the function tanh() is the hyperbolic tangent function, and the function sech() is the hyperbolic secant function;
- S211 The matrix The matrix corresponding to the mixed signal to be classified Multiply to separate each signal in the mixed signal to be classified to obtain the signal to be identified;
- the matrix The matrix corresponding to the mixed signal to be classified Multiply to get the matrix Among them, the matrix Each column of corresponds to a signal to be identified. For example, the matrix There are 3 columns, indicating that there are 3 signals to be identified.
- five high-level cumulants corresponding to each signal to be identified can be calculated: C 20 , C 21 , C 40 , C 41 , C 42 .
- the calculated high-order cumulants are usually normalized in practical applications; for example, in a specific embodiment, the formula can be used: The high-order cumulant is normalized.
- S215 Input the characteristics of the signal to be recognized into the preset support vector machine model
- the mixed signal classification method provided in the embodiment of the application does not require the classification environment. Unlike the existing mixed signal classification method, it needs to meet specific conditions before the mixed signal can be classified; therefore, the embodiment of the application provides Compared with the prior art, the classification method of mixed signals has universal applicability.
- the method provided in the embodiments of the present application also uses the preset principal component analysis method PCA to remove the influence of noise on the signal, which improves the accuracy of signal modulation recognition, for example:
- the constellation diagram of the mixed signal which is a mixture of QPSK and 16QAM, received by the receiving end, is shown in Figure 3.
- the constellation points in the constellation diagram of the mixed signal are relatively scattered. If the PCA provided in the embodiment of the application is not used to process the mixed signal, then the constellation diagram corresponding to the signal QPSK in the mixed signal is obtained, as shown in Figure 4; and the PCA provided in the embodiment of the application is used to process the mixed signal After that, the constellation diagram corresponding to the signal QPSK in the mixed signal is obtained, as shown in FIG. 5. Since a good signal is displayed as a concentrated constellation point on the constellation diagram, a noisy signal is displayed as a scattered constellation point on the constellation diagram. Therefore, comparing Figures 4 and 5, we can see that the constellation diagram shown in Figure 5 Each constellation point is more concentrated. It can be seen that using the PCA provided in the embodiment of the present application can better QPSK the signal obtained after processing the mixed signal.
- the embodiment of the present application also provides a mixed signal classification device.
- the device includes:
- the receiving module 601 is configured to receive a mixed signal, and the mixed signal contains noise and at least two different signals;
- the analysis module 602 is used to calculate the matrix corresponding to the mixed signal by using the preset principal component analysis method PCA, obtain the mixed signal to be classified and determine the number of signal types contained in the mixed signal to be classified; The mixed signal obtained after mixing the noise in the signal;
- the determining module 603 is configured to determine the separation matrix according to the number of signal types contained in the mixed signal to be classified;
- the separation module 604 is used to separate each signal in the mixed signal to be classified by using the separation matrix to obtain the signal to be identified;
- the calculation module 605 is configured to separately calculate the preset number of high-order cumulants corresponding to each signal to be identified in the signal to be identified;
- the feature module 606 is configured to use the calculated high-order cumulants as the characteristics of the signals to be identified corresponding to the high-order cumulants;
- the input module 607 is used to input the characteristics of the signal to be identified into a preset classification model; the classification model is used to calculate and output the modulation mode of the signal to be identified according to the characteristics of the signal to be identified;
- the obtaining module 608 is used to obtain the output result of the classification model; the output result includes: the modulation mode of the signal to be identified.
- the mixed signal classification device provided in the embodiment of this application has no requirements on the classification environment. Unlike the existing mixed signal classification method, it needs to meet specific conditions before the mixed signal can be classified; therefore, the embodiment of this application provides The mixed signal classification device has universal applicability compared with the prior art.
- the embodiment of the present application also provides an electronic device, as shown in FIG. 7, including a processor 701, a communication interface 702, a memory 703, and a communication bus 704.
- the processor 701, Communication interface 702, memory 703 completes mutual communication through communication bus 704,
- the memory 703 is used to store computer programs
- the processor 701 is configured to implement the steps of the mixed signal classification method provided in the embodiment of the present application when the program stored in the memory 703 is executed.
- the steps of the method for classifying mixed signals contained in the electronic equipment provided in the embodiments of the present application have no requirements for the classification environment. Unlike the existing mixed signal classification methods, which need to meet specific conditions, the mixed signals can be classified; therefore, Compared with the prior art, the electronic equipment provided by the embodiments of the present application has universal applicability.
- the communication bus mentioned in the above electronic device may be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus.
- PCI Peripheral Component Interconnect
- EISA Extended Industry Standard Architecture
- the communication bus can be divided into address bus, data bus, control bus and so on. For ease of representation, only one thick line is used in the figure, but it does not mean that there is only one bus or one type of bus.
- the communication interface is used for communication between the aforementioned electronic device and other devices.
- the memory may include random access memory (Random Access Memory, RAM), and may also include non-volatile memory (Non-Volatile Memory, NVM), such as at least one disk storage.
- NVM non-Volatile Memory
- the memory may also be at least one storage device located far away from the foregoing processor.
- the above-mentioned processor can be a general-purpose processor, including a central processing unit (CPU), a network processor (Network Processor, NP), etc.; it can also be a digital signal processor (Digital Signal Processing, DSP), a dedicated integrated Circuit (Application Specific Integrated Circuit, ASIC), Field-Programmable Gate Array (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components.
- CPU central processing unit
- NP Network Processor
- DSP Digital Signal Processing
- ASIC Application Specific Integrated Circuit
- FPGA Field-Programmable Gate Array
- a computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, it realizes any of the foregoing mixed signal Steps of classification method.
- a computer program product containing instructions is also provided, which when running on a computer, causes the computer to execute any of the mixed signal classification methods in the foregoing embodiments.
- the computer program product includes one or more computer instructions.
- the computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices.
- the computer instructions may be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions may be transmitted from a website, computer, server, or data center.
- the computer-readable storage medium may be any available medium that can be accessed by a computer or a data storage device such as a server or a data center integrated with one or more available media.
- the usable medium may be a magnetic medium (for example, a floppy disk, a hard disk, a magnetic tape), an optical medium (for example, a DVD), or a semiconductor medium (for example, a solid state disk (SSD)).
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- General Engineering & Computer Science (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Multimedia (AREA)
- Databases & Information Systems (AREA)
- Computing Systems (AREA)
- General Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Noise Elimination (AREA)
- Monitoring And Testing Of Transmission In General (AREA)
- Variable-Direction Aerials And Aerial Arrays (AREA)
- Complex Calculations (AREA)
Abstract
一种混合信号的分类方法，可以接收混合信号；利用预设的主成分分析法，对混合信号对应的矩阵进行计算，获得待分类混合信号及确定待分类混合信号中包含的信号种类个数；根据待分类混合信号中包含的信号种类个数，确定分离矩阵；利用分离矩阵分离待分类混合信号中的各类信号，得到待识别信号；分别计算待识别信号中，每一待识别信号对应的预设数量个高阶累积量；将计算得到的高阶累积量，分别作为该高阶累积量对应的待识别信号的特征；将待识别信号的特征输入预设的分类模型；获得待识别信号的调制方式。本申请实施例提供的方法，对分类环境无要求，不像现有技术，需满足特定条件，才能对混合信号分类；因此，具有普遍适用性。
Description
本申请要求于2019年4月23日提交中国专利局、申请号为201910328208.9发明名称为“混合信号的分类方法、装置及电子设备”的中国专利申请的优先权，其全部内容通过引用结合在本申请中。
本申请涉及混合信号的分类技术领域，特别是涉及一种混合信号的分类方法、装置及电子设备。
在信号的传输过程中不可避免的会混入一些干扰信号，信号接收端在接收到混入干扰信号的目标信号后，就需要对混合信号中的信号进行分类，从而，获得目标信号。由于每一种信号对应一种特定的调制方式，因此，通常是通过识别混合信号中各种信号的调制方式，对混合信号中的各种信号进行分类的。
现有的混合信号的分类方法，通常采用如下方式对混合信号中的各种信号进行分类：在混合信号中各种信号的功率相同，或目标信号的信号强度远强于混合信号中的其他信号时，先通过计算混合信号中每一种信号的信道矩阵来分离混合信号中的各种信号，再通过识别各种信号的调制方式，对混合信号中的各种信号进行分类。
然而，发明人在实现本申请的过程中发现，现有技术至少存在以下缺点：现有的混合信号的分类方法，对环境的要求比较苛刻，具体体现为：通常需要满足特定条件，才能对混合信号进行分类，该特定条件包括：混合信号中各种信号的功率相同，或目标信号的信号强度要远强于混合信号中的其他信号。可见，现有的混合信号的分类方法，在实际应用中并不具有普遍适用性，即适用场景有限。
发明内容
本申请实施例的目的在于提供一种混合信号的分类方法、装置及电子设备，以实现提供一种具有普遍适用性的混合信号分类方法。具体技术方案如 下：
为了达到上述目的，第一方面，本申请实施例提供了一种混合信号的分类方法，该方法，包括：
接收混合信号，所述混合信号中包含噪声和至少2种不同的信号；
利用预设的主成分分析法PCA，对所述混合信号对应的矩阵进行计算，获得待分类混合信号以及确定所述待分类混合信号中包含的信号种类个数；所述待分类混合信号为去除所述混合信号中的噪声后得到的混合信号；
根据所述待分类混合信号中包含的信号种类个数，确定分离矩阵；
利用所述分离矩阵分离所述待分类混合信号中的各类信号，得到待识别信号；
分别计算所述待识别信号中，每一待识别信号对应的预设数量个高阶累积量；
将计算得到的所述高阶累积量，分别作为该高阶累积量对应的待识别信号的特征；
将所述待识别信号的特征输入预设的分类模型；所述分类模型用于根据所述待识别信号的特征计算并输出所述待识别信号的调制方式；
获得所述分类模型的输出结果；所述输出结果中包括：所述待识别信号的调制方式。
可选的，利用预设的主成分分析法PCA，对所述混合信号对应的矩阵进行计算，获得待分类混合信号以及确定所述待分类混合信号中包含的信号种类个数的步骤，包括：
对所述混合信号对应的矩阵R，进行归一化处理，计算获得矩阵
所述矩阵
为所述混合信号对应的矩阵R，进行归一化处理得到的矩阵；
对所述矩阵
进行中心化处理，使所述矩阵
的平均值为0，计算获得矩阵
计算所述矩阵
的自相关矩阵；并对所述矩阵
的自相关矩阵
进行奇异值分解，得到
其中，
为矩阵
的转置共轭矩阵，
为
的转置共轭矩阵，
为正交矩阵，
是矩阵
的第n个列，对角矩阵
为
N为接收所述混合信号的天线个数；λ
1,…,λ
N为所述自相关矩阵
的奇异值；
将所述奇异值λ
1,…,λ
N由小到大排列；并将所述奇异值λ
1,…,λ
N中数值小于预设阈值的奇异值的数值设为0，计算获得对角矩阵
令
利用预设的公式
计算获得所述待分类混合信号；所述待分类混合信号对应的矩阵为
根据所述奇异值λ
1,…,λ
N中，不为0的奇异值的个数，确定所述待分类混合信号中包含的信号种类个数。
可选的，根据所述待分类混合信号中包含的信号种类个数，确定分离矩阵的步骤，包括：
获得预设的迭代次数和初始分离矩阵w；
根据所述迭代次数和所述初始分离矩阵w，利用预设的公式
计算获得分离矩阵w；其中，所述分离矩阵w有M行N列，M为所述待分类混合信号中包含的信号种类个数，N为接收所述混合信号的天线个数，L为信号接收次数，
表示所述分离 矩阵w的第m列在k次迭代后的结果，
向量
为所述待分类混合信号
的第i列。
可选的，所述利用所述分离矩阵分离所述待分类混合信号中的各个信号，得到待识别信号的步骤，包括：
对所述分离矩阵w，进行归一化处理，得到矩阵
将所述分离矩阵
与所述待分类混合信号对应的矩阵
相乘，分离所述待分类混合信号中的各个信号，得到待识别信号。
可选的，所述将计算得到的所述高阶累积量，分别作为该高阶累积量对应的待识别信号的特征的步骤，包括：
对所述高阶累积量进行归一化处理；
将归一化处理后的高阶累积量，作为所述高阶累积量对应的待识别信号的特征。
可选的，所述分类模型为支持向量机模型；
在所述将所述待识别信号的特征输入预设的分类模型的步骤之前，还包括：
将带有标签的训练样本，输入当前支持向量机模型中，获得当前支持向量机模型输出的各个训练样本对应的调制方式；
根据所述当前支持向量机模型输出结果和所述训练样本的标签，使用预设的损失函数计算损失值；
根据所述损失值，调整所述当前支持向量机中的参数，得到所述支持向量机模型。
第二方面，本申请实施例提供了一种混合信号的分类装置，该装置包括：
接收模块，用于接收混合信号，所述混合信号中包含噪声和至少2种不同的信号；
分析模块，用于利用预设的主成分分析法PCA，对所述混合信号对应的 矩阵进行计算，获得待分类混合信号以及确定所述待分类混合信号中包含的信号种类个数；所述待分类混合信号为去除所述混合信号中的噪声后得到的混合信号；
确定模块，用于根据所述待分类混合信号中包含的信号种类个数，确定分离矩阵；
分离模块，用于利用所述分离矩阵分离所述待分类混合信号中的各个信号，得到待识别信号；
计算模块，用于分别计算所述待识别信号中，每一待识别信号对应的预设数量个高阶累积量；
特征模块，用于将计算得到的所述高阶累积量，分别作为该高阶累积量对应的待识别信号的特征；
输入模块，用于将所述待识别信号的特征输入预设的分类模型；所述分类模型用于根据所述待识别信号的特征计算并输出所述待识别信号的调制方式；
获得模块，用于获得所述分类模型的输出结果；所述输出结果中包括：所述待识别信号的调制方式。
第三方面，本申请实施例提供了一种电子设备，包括：处理器、通信接口、存储器和通信总线，其中，处理器，通信接口，存储器通过通信总线完成相互间的通信；
存储器，用于存放计算机程序；
处理器，用于执行存储器上所存放的程序时，实现如下步骤：
接收混合信号，所述混合信号中包含噪声和至少2种不同的信号；
利用预设的主成分分析法PCA，对所述混合信号对应的矩阵进行计算，获得待分类混合信号以及确定所述待分类混合信号中包含的信号种类个数；所述待分类混合信号为去除所述混合信号中的噪声后得到的混合信号；
根据所述待分类混合信号中包含的信号种类个数，确定分离矩阵；
利用所述分离矩阵分离所述待分类混合信号中的各类信号，得到待识别信号；
分别计算所述待识别信号中，每一待识别信号对应的预设数量个高阶累积量；
将计算得到的所述高阶累积量，分别作为该高阶累积量对应的待识别信号的特征；
将所述待识别信号的特征输入预设的分类模型；所述分类模型用于根据所述待识别信号的特征计算并输出所述待识别信号的调制方式；
获得所述分类模型的输出结果；所述输出结果中包括：所述待识别信号的调制方式。
第四方面，本申请实施例提供了一种计算机可读存储介质，所述计算机可读存储介质内存储有计算机程序，所述计算机程序被处理器执行时实现如下步骤：
接收混合信号，所述混合信号中包含噪声和至少2种不同的信号；
利用预设的主成分分析法PCA，对所述混合信号对应的矩阵进行计算，获得待分类混合信号以及确定所述待分类混合信号中包含的信号种类个数；所述待分类混合信号为去除所述混合信号中的噪声后得到的混合信号；
根据所述待分类混合信号中包含的信号种类个数，确定分离矩阵；
利用所述分离矩阵分离所述待分类混合信号中的各类信号，得到待识别信号；
分别计算所述待识别信号中，每一待识别信号对应的预设数量个高阶累积量；
将计算得到的所述高阶累积量，分别作为该高阶累积量对应的待识别信号的特征；
将所述待识别信号的特征输入预设的分类模型；所述分类模型用于根据所述待识别信号的特征计算并输出所述待识别信号的调制方式； 获得所述分类模型的输出结果；所述输出结果中包括：所述待识别信号的调制方式。
本申请实施例提供的一种混合信号的分类方法、装置及电子设备，可以实现在接收到混合信号后，利用预设的主成分分析法PCA，对所述混合信号对应的矩阵进行计算，获得待分类混合信号以及确定所述待分类混合信号中包含的信号种类个数；根据所述待分类混合信号中包含的信号种类个数，确定分离矩阵；利用所述分离矩阵分离所述待分类混合信号中的各个信号，得到待识别信号；分别计算所述待识别信号中，每一待识别信号对应的预设数量个高阶累积量；将计算得到的所述高阶累积量，分别作为该高阶累积量对应的待识别信号的特征；将所述待识别信号的特征输入预设的分类模型；获得所述分类模型的输出结果。本申请实施例提供的混合信号的分类方法，对分类环境没有要求，不像现有的混合信号的分类方法，需要满足特定的条件，才能对混合信号进行分类；因此，本申请实施例提供的混合信号的分类方法，相较于现有技术具有普遍适用性。
当然，实施本申请的任一产品或方法并不一定需要同时达到以上所述的所有优点。
为了更清楚地说明本申请实施例和现有技术的技术方案，下面对实施例和现有技术中所需要使用的附图作简单地介绍，显而易见地，下面描述中的附图仅仅是本申请的一些实施例，对于本领域普通技术人员来讲，在不付出创造性劳动的前提下，还可以根据这些附图获得其他的附图。
图1a为本申请实施例提供的混合信号的分类方法的流程示意图；
图1b为信号发送和接收的一种模型示意图；
图2为本申请实施例提供的混合信号的分类方法的另一流程示意图；
图3为接收端接收到的由信号QPSK和16QAM混合而成的混合信号的星座图；
图4为不采用本申请实施例提供的PCA对混合信号进行处理得到混合信号中信号QPSK对应的星座图；
图5为利用本申请实施例提供的PCA对混合信号处理之后得到混合信号中信号QPSK对应的星座图；
图6为本申请实施例提供的混合信号的分类装置的结构示意图；
图7为本申请实施例提供的一种电子设备的结构示意图。
为使本申请的目的、技术方案、及优点更加清楚明白，以下参照附图并举实施例，对本申请进一步详细说明。显然，所描述的实施例仅仅是本申请一部分实施例，而不是全部的实施例。基于本申请中的实施例，本领域普通技术人员在没有作出创造性劳动前提下所获得的所有其他实施例，都属于本申请保护的范围。
为了解决现有技术对分类环境要求较高，需要满足特定的条件，才能对混合信号进行分类的问题，本申请实施例提供了一种混合信号的分类方法、装置及电子设备。
下面首先对本申请实施例所提供的一种混合信号的分类方法进行介绍。其中，本申请实施例所提供的一种混合信号的分类方法可以应用于信号接收端，当然并不局限于此。并且，本申请所针对的混合信号为对一种信号混有一些干扰信号后所得到的、包含多种不同信号的信号数据。可以理解的是，该干扰信号不同于信号传输过程中噪声，通过该干扰信号可以使得传输的一种信号由一种变更为多种。
本申请实施例提供了一种混合信号的分类方法。参见图1a，该方法包括：
S101：接收混合信号，混合信号中包含噪声和至少2种不同的信号；
具体的，由于一个信号源对应一种信号，比如，有3个信号源，就表示同时发送3种信号，参见图1b，可以用M个信号源同时发送L次信号，用N根天线接收这L次信号，为了保证混合信号能够被完全接收，可以让M≤N，其中，L可以为不小于1的正整数。
那么，第l次接收信号可以表示为r
l＝[r
l,1,…,r
l,N]
T(l＝1,…,L)，则r
l可以表示 为r
l＝Hs
l+v
l，其中，s
l＝[s
l,1,…,s
l,M]
T是第l次发送信号，
是信道矩阵，矩阵H的第(n,m)个符号h
nm被定义为复信道第m次发送信号和第n次接收信号之间的信道增益，噪声
为复数正态分布，I是单位矩阵。
需要说明的是，所接收到的混合信号为在一种信号传输过程中混入其他种信号后所得到的信号，也即在一个信号源的信号的传输过程中混入其他信号源的信号所得到信号。
所谓的复信道是传输复信号的信道。其中，该复信号是一种信号表征方法，具体而言：利用复数的模值表示信号的幅度，并利用复数的幅角表示信号的相位。
S102：利用预设的主成分分析法PCA，对混合信号对应的矩阵进行计算，获得待分类混合信号以及确定待分类混合信号中包含的信号种类个数；待分类混合信号为去除混合信号中的噪声后得到的混合信号；
其中，主成分分析(Principal Component Analysis，PCA)，是一种统计方法。通过正交变换将一组可能存在相关性的变量转换为一组线性不相关的变量，转换后的这组变量叫主成分。本领域技术人员可以理解的是，PCA可以用作噪音的过滤，因为任何一个成分的变化影响都远远大于随机噪声的影响的，所以针对于噪声，各种的成分相对不受影响，即可以用主成分来重构原始数据。关于利用预设的主成分分析法PCA，对混合信号对应的矩阵进行计算，获得待分类混合信号以及确定待分类混合信号中包含的信号种类个数的具体实现方式可以与现有技术相同，本申请在此不作详述。
S103：根据待分类混合信号中包含的信号种类个数，确定分离矩阵；
示例性地，在一种实现方式中，根据待分类混合信号中包含的信号种类个数，确定分离矩阵的步骤，可以包括：
获得预设的迭代次数和初始分离矩阵w；
根据迭代次数和初始分离矩阵w，利用预设的公式
计算获得分离矩阵w；其中，分离矩阵w有M行N列，M为待分类混合信号中包含的信号种类个数，N为接收混合信号的天线个数，L为信号接收次数，
表示分离矩阵w的第m列在k次迭代后的结果，
向量
为待分类混合信号
的第i列。可以理解的是，L也可以称为信号采样次数，并且，L的数值与信号发送次数相同。
S104：利用分离矩阵分离待分类混合信号中的各个信号，得到待识别信号；
需要强调的是，利用分离矩阵分离待分类混合信号中的各个信号，得到待识别信号，即为利用分离矩阵分离待分类混合信号中的各种信号，得到多种待识别信号。
示例性地，利用分离矩阵分离待分类混合信号中的各个信号，得到待识别信号的步骤，可以包括：
对分离矩阵w，进行归一化处理，得到矩阵
将矩阵
与待分类混合信号对应的矩阵
相乘，分离待分类混合信号中的各个信号，得到待识别信号。
S105：分别计算待识别信号中，每一待识别信号对应的预设数量个高阶累积量；
在实际应用中，为了提高分类结果的准确性，可以针对每一待识别信号，分别计算该待识别信号对应的多个高阶累积量，将计算得到的多个高阶累积量，都作为该待识别信号的特征。可以理解的是，每一种待识别信号均有预设数量个高阶累积量，其中，计算每一种待识别信号对应的预设数量个高阶累计量的过程可以参照现有技术中任一种计算信号的高阶累积量的实现方式，在此不作赘述。
S106：将计算得到的高阶累积量，分别作为该高阶累积量对应的待识别 信号的特征；
可以理解的是，每一种待识别信号均对应由高阶累积量，那么，针对每一种待识别信号，可以将该种待识别信号对应的高阶累积量作为该种待识别信号对应的特征。
示例性地，在一种实现方式中，将计算得到的高阶累积量，分别作为该高阶累积量对应的待识别信号的特征的步骤，可以包括：
对高阶累积量进行归一化处理；
将归一化处理后的高阶累积量，作为高阶累积量对应的待识别信号的特征。
具体而言，针对每一种待识别信号，将该种待识别信号对应的高阶累积量进行归一化处理，将归一化处理后的高阶累积量，作为该种待识别信号的特征。
示例性地，在另一种实现方式中，针对每一种待识别信号，可以将该种待识别信号对应地高阶累积量，确定为该种待识别信号的特征。
S107：将待识别信号的特征输入预设的分类模型；分类模型用于根据待识别信号的特征计算并输出待识别信号的调制方式；
其中，该分类模型的种类可以为多种，例如：神经网络模型，或者，支持向量机模型，等等。
S108：获得分类模型的输出结果；输出结果中包括：待识别信号的调制方式。
在得到各种待识别信号的调制方式后，即可完成对混合信号中各种信号的分类。
由图1a所示的实施例可见，本申请实施例提供的混合信号的分类方法，对分类环境没有要求，不像现有的混合信号的分类方法，需要满足特定的条件，才能对混合信号进行分类；因此，本申请实施例提供的混合信号的分类方法，相较于现有技术具有普遍适用性。
在一种具体的实施例中，分类模型可以为支持向量机模型；
在将待识别信号的特征输入预设的分类模型的步骤之前，还包括：
将带有标签的训练样本，输入当前支持向量机模型中，获得当前支持向量机模型输出的各个训练样本对应的调制方式；
根据当前支持向量机模型输出结果和训练样本的标签，使用预设的损失函数计算损失值；
根据损失值，调整当前支持向量机中的参数，得到支持向量机模型。
其中，训练样本为信号数据，训练样本的标签信号数据的调制方式。
例如：支持向量机模型可以为
其中参数
表示权重，参数
表示偏置，
为信号的特征。
设有S个训练样本
该训练样本对应的特征向量分别为
训练样本对应的标签分别为ξ
1,ξ
2,...,ξ
s，将特征向量
及其对应的标签ξ
1,ξ
2,...,ξ
s，输入当前支持向量机模型，获得当前支持向量机模型输出的各个训练样本对应的调制方式；根据当前支持向量机模型输出结果和训练样本的标签，使用预设的损失函数：
计算损失值；根据损失值，调整当前支持向量机中的参数
和
得到支持向量机模型。其中，t
s∈{1,…,I}表示训练样本
属于第t
s种调制方式，即训练样本
属于第t
s种信号；I代表了该支持向量机模型能识别出多少种调制方式，例如，I为3，则表明该支持向量机模型能识别出3种调制方式；i∈{1,…,I}\{t
s}表示松弛变量；
为惩罚项的系数，是一个常数，用于控制对错分样本的惩罚程度；
表示输出结果和实际结果之间的误差值。
在一种具体的实施例中，可以将松弛变量和惩罚项的系数分别取值为0.1和0.2。
在一种具体的实施例中，可以利用公式
根据调制方式计算得到，该与调制方式对应的特征。其中，m
test表示调制方式；
表示与m
test对应的特征；函数argmax()用于计算，最大结果值对应的参数t的值，也就是得到最大可能调制方式，对应的特征。
在实际应用中，本申请实施例提供的方法，总的思想流程可以为：
第一步，利用预设的主成分分析法PCA去除混合信号中的噪声避免噪声对分类结果的影响，获得待分类混合信号对应的矩阵
第二步，根据待分类混合信号中包含的信号种类个数，确定分离矩阵w，用分离矩阵w与第一步得到的待分类混合信号对应的矩阵
相乘，降低待分类混合信号中的各种类信号间的相关性，得到待识别信号。在实际应用中为了减少计算量，通常在分离矩阵w与待分类混合信号
相乘之前，会先对分离矩阵w进行归一化处理，得到矩阵
再用矩阵
与待分类混合信号对应的矩阵
相乘。
第三步，针对每一待识别信号，计算该待识别信号对应的多个高阶累积量，将计算得到的高阶累积量作为该待识别信号的特征，将待识别信号的特征输入预设的分类模型：支持向量机模型，通过支持向量机模型对待识别信号的调制格式进行识别，从而达到对混合信号进行分类的目的。
参见图2，以下列举一个实际的例子，对本申请实施例提供的方法，作进一步的说明：
S201：用N根天线接收混合信号；其中，N为大于1的正整数；
在实际应用中，为了保证信号能被完全接收到，通常会使天线的数量大于或等于混合信号中可能包含的信号个数。
S202：对该混合信号对应的矩阵R，进行归一化处理，计算获得矩阵
S203：对矩阵
进行中心化处理，使矩阵
的平均值为0，计算获得矩阵
S204：计算矩阵
的自相关矩阵；并对矩阵
的自相关矩阵
进行奇异值分解，得到
其中，
为矩阵
的转置共轭矩阵，
为
的转置共轭矩阵，
为正交矩阵，
是矩阵
的第n个列，对角矩阵
为
N为接收混合信号的天线个数；λ
1,…,λ
N为自相关矩阵
的奇异值；
S205：将奇异值λ
1,…,λ
N由小到大排列；并将奇异值λ
1,…,λ
N中数值小于预设阈值的奇异值的数值设为0，计算获得对角矩阵
S206：令
利用预设的公式
计算获得待分类混合信号；待分类混合信号对应的矩阵为
S207：根据奇异值λ
1,…,λ
N中，不为0的奇异值的个数，确定待分类混合信号中包含的信号种类个数；
例如：奇异值λ
1,…,λ
N中，不为0的奇异值的个数为3个，这样，就确定了待分类混合信号中包含3种不同的信号。
S208：获得预设的迭代次数和初始分离矩阵；
具体的，迭代次数和初始分离矩阵均可以人为设定，比如，将迭代次数设为10次，将初始分离矩阵设为有M行N列单位矩阵，其中，M为接收到的混合信号中包含的信号种类数，N为天线数。例如，接收到的混合信号中包含的信号种类数为3，天线数为10，则初始分离矩阵为3行10列的单位矩阵。
S209：根据迭代次数和初始分离矩阵，利用预设的公式
计算获得分离矩阵w；其中，分离矩阵w有M行N列，M为待分类混合信号中包含的信号种类个数，N为接收混合信号的天线个数，L为信号接收次数，
表示分离矩阵w的第m列在k次迭代后的结果，
向量
为待分类混合信号
的第i列，函数tanh()为双曲正切函数，函数sech()为双曲正割函数；
S210：对分离矩阵w，进行归一化处理，得到矩阵
S211：将矩阵
与待分类混合信号对应的矩阵
相乘，分离待分类混合信号中的各个信号，得到待识别信号；
具体的，就是让矩阵
与待分类混合信号对应的矩阵
相乘，使得矩阵
中各列不相关。
具体的，将矩阵
与待分类混合信号对应的矩阵
相乘，得到矩阵
其中，矩阵
的每一列对应一种待识别信号。比如，矩阵
有3列，则表明有3种待识别信号。
S212：分别计算待识别信号中，每一待识别信号对应的预设数量个高阶累积量；
在一种具体的实施例中，可以计算每一待识别信号对应的5个高阶累积量：C
20、C
21、C
40、C
41、C
42。
高阶累积量的计算方法是现有技术，下面仅给出现有技术中不同的 高阶累积量的计算公式：
对于随机变量x＝[x
1，…，x
n]
T，n＝1，2，…,N，其不同的高阶累积量的计算公式可以为：
根据上述公式可以分别计算得到待识别信号所选择的多个高阶累积量。
S213：对高阶累积量进行归一化处理；
为了简化计算，在实际应用中通常会对计算得到的高阶累积量进行归一化处理；例如，在一种具体的实施例中可以利用公式：
高阶累积量进行归一化处理。
S214：将归一化处理后的高阶累积量，作为高阶累积量对应的待识别信号的特征；
S215：将待识别信号的特征输入预设的支持向量机模型；
S216：获得该支持向量机模型的输出结果。
本申请实施例提供的混合信号的分类方法，对分类环境没有要求，不像现有的混合信号的分类方法，需要满足特定的条件，才能对混合信号进行分类；因此，本申请实施例提供的混合信号的分类方法，相较于现有技术具有普遍适用性。
另外，本申请实施例提供的方法，还利用预设的主成分分析法PCA去除噪声对信号的影响，提高了信号调制方式识别的准确率，例如：
接收端接收到的，由信号QPSK和16QAM混合而成的混合信号的星座图，如图3所示。参照图3所示，由于存在噪声，混合信号的星座图中的星座点较为分散。如果不利用本申请实施例提供的PCA，对混合信号进行处理，那么，得到混合信号中信号QPSK对应的星座图，如图4所示；而利用本申请实施例提供的PCA，对混合信号处理之后，得到混合信号中信号QPSK对应的星座图，如图5所示。由于良好的信号在星座图上显示是很集中的星座点，有噪声的信号在星座图上显示是分散的星座点，因此，对比图4和图5可知，图5所示的星座图中的各个星座点更为集中，可见，利用本申请实施例提供的PCA，对混合信号处理之后所得到的信号QPSK更为良好。
综上可知，经过了PCA处理之后，提高了星座图图像的清晰度，从而也就提高了信号调制方式识别的准确率。
与图1a所示实施例对应的，本申请实施例还提供了一种混合信号的分类装置，参见图6，该装置包括：
接收模块601，用于接收混合信号，混合信号中包含噪声和至少2种不同的信号；
分析模块602，用于利用预设的主成分分析法PCA，对混合信号对应的矩阵进行计算，获得待分类混合信号以及确定待分类混合信号中包含的信号种类个数；待分类混合信号为去除混合信号中的噪声后得到的混合信号；
确定模块603，用于根据待分类混合信号中包含的信号种类个数，确定分离矩阵；
分离模块604，用于利用分离矩阵分离待分类混合信号中的各个信号，得到待识别信号；
计算模块605，用于分别计算待识别信号中，每一待识别信号对应的预设数量个高阶累积量；
特征模块606，用于将计算得到的高阶累积量，分别作为该高阶累积量对应的待识别信号的特征；
输入模块607，用于将待识别信号的特征输入预设的分类模型；分类模型用于根据待识别信号的特征计算并输出待识别信号的调制方式；
获得模块608，用于获得分类模型的输出结果；输出结果中包括：待识别信号的调制方式。
本申请实施例提供的混合信号的分类装置，对分类环境没有要求，不像现有的混合信号的分类方法，需要满足特定的条件，才能对混合信号进行分类；因此，本申请实施例提供的混合信号的分类装置，相较于现有技术具有普遍适用性。
与图1a所示实施例对应的，本申请实施例还提供了一种电子设备，如图7所示，包括处理器701、通信接口702、存储器703和通信总线704，其中，处理器701，通信接口702，存储器703通过通信总线704完成相互间的通信，
存储器703，用于存放计算机程序；
处理器701，用于执行存储器703上所存放的程序时，实现本申请实施例所提供的混合信号的分类方法的步骤。
本申请实施例提供的电子设备中包含的混合信号的分类方法步骤，对分类环境没有要求，不像现有的混合信号的分类方法，需要满足特定的条件，才能对混合信号进行分类；因此，本申请实施例提供的电子设备，相较于现有技术具有普遍适用性。
上述电子设备提到的通信总线可以是外设部件互连标准(Peripheral Component Interconnect，PCI)总线或扩展工业标准结构(Extended Industry Standard Architecture，EISA)总线等。该通信总线可以分为地址总线、数据总线、控制总线等。为便于表示，图中仅用一条粗线表示，但并不表示仅有一根总线或一种类型的总线。
通信接口用于上述电子设备与其他设备之间的通信。
存储器可以包括随机存取存储器(Random Access Memory，RAM)，也可以包括非易失性存储器(Non-Volatile Memory，NVM)，例如至少一个磁盘存储器。可选的，存储器还可以是至少一个位于远离前述处理器的存储装置。
上述的处理器可以是通用处理器，包括中央处理器(Central Processing Unit，CPU)、网络处理器(Network Processor，NP)等；还可以是数字信号处理器(Digital Signal Processing，DSP)、专用集成电路(Application Specific Integrated Circuit，ASIC)、现场可编程门阵列(Field-Programmable Gate Array，FPGA)或者其他可编程逻辑器件、分立门或者晶体管逻辑器件、分立硬件组件。
在本申请提供的又一实施例中，还提供了一种计算机可读存储介质，该计算机可读存储介质内存储有计算机程序，所述计算机程序被处理器执行时实现上述任一混合信号的分类方法的步骤。
在本申请提供的又一实施例中，还提供了一种包含指令的计算机程序产品，当其在计算机上运行时，使得计算机执行上述实施例中任一混合信号的分类方法。
在上述实施例中，可以全部或部分地通过软件、硬件、固件或者其任意组合来实现。当使用软件实现时，可以全部或部分地以计算机程序产品的形式实现。所述计算机程序产品包括一个或多个计算机指令。在计算机上加载和执行所述计算机程序指令时，全部或部分地产生按照本申请实施例所述的流程或功能。所述计算机可以是通用计算机、专用计算机、计算机网络、或者其他可编程装置。所述计算机指令可以存储在计算机可读存储介质中，或者从一个计算机可读存储介质向另一个计算机可读存储介质传输，例如，所述计算机指令可以从一个网站站点、计算机、服务器或数据中心通过有线(例如同轴电缆、光纤、数字用户线(DSL))或无线(例如红外、无线、微波等)方式向另一个网站站点、计算机、服务器或数据中心进行传输。所述计算机可读存储介质可以是计算机能够存取的任何可用介质或者是包含一个或多个可用介质集成的服务器、数据中心等数据存储设备。所述可用介质可以是磁性介质，(例如，软盘、硬盘、磁带)、光介质(例如，DVD)、或者半导体介质(例如固态硬盘Solid State Disk(SSD))等。
需要说明的是，在本文中，诸如第一和第二等之类的关系术语仅仅用来将一个实体或者操作与另一个实体或操作区分开来，而不一定要求或者暗示这些实体或操作之间存在任何这种实际的关系或者顺序。而且，术语“包括”、 “包含”或者其任何其他变体意在涵盖非排他性的包含，从而使得包括一系列要素的过程、方法、物品或者设备不仅包括那些要素，而且还包括没有明确列出的其他要素，或者是还包括为这种过程、方法、物品或者设备所固有的要素。在没有更多限制的情况下，由语句“包括一个……”限定的要素，并不排除在包括所述要素的过程、方法、物品或者设备中还存在另外的相同要素。
本说明书中的各个实施例均采用相关的方式描述，各个实施例之间相同相似的部分互相参见即可，每个实施例重点说明的都是与其他实施例的不同之处。尤其，对于装置实施例而言，由于其基本相似于方法实施例，所以描述的比较简单，相关之处参见方法实施例的部分说明即可。
以上所述仅为本申请的较佳实施例而已，并不用以限制本申请，凡在本申请的精神和原则之内，所做的任何修改、等同替换、改进等，均应包含在本申请保护的范围之内。
Claims (9)
- 一种混合信号的分类方法，其特征在于，包括：接收混合信号，所述混合信号中包含噪声和至少2种不同的信号；利用预设的主成分分析法PCA，对所述混合信号对应的矩阵进行计算，获得待分类混合信号以及确定所述待分类混合信号中包含的信号种类个数；所述待分类混合信号为去除所述混合信号中的噪声后得到的混合信号；根据所述待分类混合信号中包含的信号种类个数，确定分离矩阵；利用所述分离矩阵分离所述待分类混合信号中的各类信号，得到待识别信号；分别计算所述待识别信号中，每一待识别信号对应的预设数量个高阶累积量；将计算得到的所述高阶累积量，分别作为该高阶累积量对应的待识别信号的特征；将所述待识别信号的特征输入预设的分类模型；所述分类模型用于根据所述待识别信号的特征计算并输出所述待识别信号的调制方式；获得所述分类模型的输出结果；所述输出结果中包括：所述待识别信号的调制方式。
- 根据权利要求1所述的方法，其特征在于，利用预设的主成分分析法PCA，对所述混合信号对应的矩阵进行计算，获得待分类混合信号以及确定所述待分类混合信号中包含的信号种类个数的步骤，包括：对所述混合信号对应的矩阵R，进行归一化处理，计算获得矩阵所述矩阵
为所述混合信号对应的矩阵R，进行归一化处理得到的矩阵；
对所述矩阵进行中心化处理，使所述矩阵
的平均值为0，计算获得矩阵
计算所述矩阵的自相关矩阵；并对所述矩阵
的自相关矩阵
进行奇异值分解，得到
其中，
为矩阵
的转置共轭矩阵，
为
的转置共轭矩阵，
为正交矩阵，
是矩阵
的第n个列，对角矩阵
为
N为接收所述混合信号的天线个数；λ 1,…,λ N为所述自相关矩阵
的奇异值；
将所述奇异值λ 1,…,λ N由小到大排列；并将所述奇异值λ 1,…,λ N中数值小于预设阈值的奇异值的数值设为0，计算获得对角矩阵
令利用预设的公式
计算获得所述待分类混合信号；所述待分类混合信号对应的矩阵为
根据所述奇异值λ 1,…,λ N中，不为0的奇异值的个数，确定所述待分类混合信号中包含的信号种类个数。 - 根据权利要求2所述的方法，其特征在于，根据所述待分类混合信号中包含的信号种类个数，确定分离矩阵的步骤，包括：获得预设的迭代次数和初始分离矩阵w；根据所述迭代次数和所述初始分离矩阵w，利用预设的公式计算获得分离矩阵w；其中，所述分离矩阵w有M行N列，M为所述待分类混合信号中包含的信号种类个数，N为接收所述混合信号的天线个数，L为信号接收次数，
表示所述分离 矩阵w的第m列在k次迭代后的结果，
向量
为所述待分类混合信号
的第i列。
- 根据权利要求3所述的方法，其特征在于，所述利用所述分离矩阵分离所述待分类混合信号中的各个信号，得到待识别信号的步骤，包括：对所述分离矩阵w，进行归一化处理，得到矩阵
将所述分离矩阵与所述待分类混合信号对应的矩阵
相乘，分离所述待分类混合信号中的各个信号，得到待识别信号。
- 根据权利要求1所述的方法，其特征在于，所述将计算得到的所述高阶累积量，分别作为该高阶累积量对应的待识别信号的特征的步骤，包括：对所述高阶累积量进行归一化处理；将归一化处理后的高阶累积量，作为所述高阶累积量对应的待识别信号的特征。
- 根据权利要求1所述的方法，其特征在于，所述分类模型为支持向量机模型；在所述将所述待识别信号的特征输入预设的分类模型的步骤之前，还包括：将带有标签的训练样本，输入当前支持向量机模型中，获得当前支持向量机模型输出的各个训练样本对应的调制方式；根据所述当前支持向量机模型输出结果和所述训练样本的标签，使用预设的损失函数计算损失值；根据所述损失值，调整所述当前支持向量机中的参数，得到所述支持向量机模型。
- 一种混合信号的分类装置，其特征在于，所述装置包括：接收模块，用于接收混合信号，所述混合信号中包含噪声和至少2种不同的信号；分析模块，用于利用预设的主成分分析法PCA，对所述混合信号对应的矩阵进行计算，获得待分类混合信号以及确定所述待分类混合信号中包含的信号种类个数；所述待分类混合信号为去除所述混合信号中的噪声后得到的混合信号；确定模块，用于根据所述待分类混合信号中包含的信号种类个数，确定分离矩阵；分离模块，用于利用所述分离矩阵分离所述待分类混合信号中的各个信号，得到待识别信号；计算模块，用于分别计算所述待识别信号中，每一待识别信号对应的预设数量个高阶累积量；特征模块，用于将计算得到的所述高阶累积量，分别作为该高阶累积量对应的待识别信号的特征；输入模块，用于将所述待识别信号的特征输入预设的分类模型；所述分类模型用于根据所述待识别信号的特征计算并输出所述待识别信号的调制方式；获得模块，用于获得所述分类模型的输出结果；所述输出结果中包括：所述待识别信号的调制方式。
- 一种电子设备，其特征在于，包括处理器、通信接口、存储器和通信总线，其中，处理器，通信接口，存储器通过通信总线完成相互间的通信；存储器，用于存放计算机程序；处理器，用于执行存储器上所存放的程序时，实现权利要求1-6任一所述的方法步骤。
- 一种计算机可读存储介质，其特征在于，所述计算机可读存储介质内存储有计算机程序，所述计算机程序被处理器执行时实现权利要求1-6任一所述的方法步骤。
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| US17/602,692 US11816180B2 (en) | 2019-04-23 | 2020-03-10 | Method and apparatus for classifying mixed signals, and electronic device |
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910328208.9 | 2019-04-23 | ||
| CN201910328208.9A CN110059757B (zh) | 2019-04-23 | 2019-04-23 | 混合信号的分类方法、装置及电子设备 |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| WO2020215911A1 true WO2020215911A1 (zh) | 2020-10-29 |
Family
ID=67320292
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| PCT/CN2020/078573 WO2020215911A1 (zh) | 2019-04-23 | 2020-03-10 | 混合信号的分类方法、装置及电子设备 |
Country Status (3)
| Country | Link |
|---|---|
| US (1) | US11816180B2 (zh) |
| CN (1) | CN110059757B (zh) |
| WO (1) | WO2020215911A1 (zh) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110059757B (zh) | 2019-04-23 | 2021-04-09 | 北京邮电大学 | 混合信号的分类方法、装置及电子设备 |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103973411A (zh) * | 2014-04-10 | 2014-08-06 | 沈阳理工大学 | 一种差分空时编码信号盲检测方法 |
| CN105721368A (zh) * | 2016-01-25 | 2016-06-29 | 北京邮电大学 | 一种混叠数字信号识别方法和装置 |
| CN106487730A (zh) * | 2016-10-19 | 2017-03-08 | 北京邮电大学 | 一种盲信号调制识别的方法及装置 |
| CN110059757A (zh) * | 2019-04-23 | 2019-07-26 | 北京邮电大学 | 混合信号的分类方法、装置及电子设备 |
Family Cites Families (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP4449871B2 (ja) * | 2005-01-26 | 2010-04-14 | ソニー株式会社 | 音声信号分離装置及び方法 |
| US8170152B2 (en) * | 2007-07-12 | 2012-05-01 | Bae Systems Information And Electronic Systems Integration Inc. | Method and apparatus for multiple signal identification and finding the basis functions of the received signal |
| CN102445650B (zh) * | 2011-09-22 | 2014-09-24 | 重庆大学 | 基于盲信号分离算法的电路故障诊断方法 |
| CN102647259B (zh) * | 2012-04-07 | 2014-09-17 | 西安电子科技大学 | 基于奇异值分解的多用户mimo系统上行链路传输方法 |
| CN104144020B (zh) * | 2014-06-30 | 2016-08-24 | 电子科技大学 | 一种信号方向未知的天线阵列相位响应参数的测定方法 |
| CN104181513B (zh) * | 2014-07-30 | 2016-08-24 | 西安电子科技大学 | 一种雷达天线阵元位置的校正方法 |
| US10102320B2 (en) * | 2015-02-26 | 2018-10-16 | Autodesk, Inc. | Predictive multi-user client-server electronic circuit design system utilizing machine learning techniques |
| CN104757966A (zh) * | 2015-04-13 | 2015-07-08 | 深圳市飞马与星月科技研究有限公司 | 胎儿心电信号的提取方法及装置 |
| CN106842195B (zh) * | 2016-12-26 | 2020-12-25 | 中国科学院合肥物质科学研究院 | 基于计算关联成像的多物体同时成像及加密方法 |
| CN107276938A (zh) * | 2017-06-28 | 2017-10-20 | 北京邮电大学 | 一种数字信号调制方式识别方法及装置 |
| CN109274625B (zh) * | 2018-11-12 | 2020-06-19 | 北京邮电大学 | 一种信息调制方式确定方法、装置、电子设备及存储介质 |
-
2019
- 2019-04-23 CN CN201910328208.9A patent/CN110059757B/zh active Active
-
2020
- 2020-03-10 US US17/602,692 patent/US11816180B2/en active Active
- 2020-03-10 WO PCT/CN2020/078573 patent/WO2020215911A1/zh active Application Filing
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103973411A (zh) * | 2014-04-10 | 2014-08-06 | 沈阳理工大学 | 一种差分空时编码信号盲检测方法 |
| CN105721368A (zh) * | 2016-01-25 | 2016-06-29 | 北京邮电大学 | 一种混叠数字信号识别方法和装置 |
| CN106487730A (zh) * | 2016-10-19 | 2017-03-08 | 北京邮电大学 | 一种盲信号调制识别的方法及装置 |
| CN110059757A (zh) * | 2019-04-23 | 2019-07-26 | 北京邮电大学 | 混合信号的分类方法、装置及电子设备 |
Non-Patent Citations (1)
| Title |
|---|
| ZHANG MINGMING: "The Research of Typical Wireless Communication Signal and Biomedical Signal on Technology of Pattern Recognition", CHINESE MASTER’S THESES FULL-TEXT DATABASE, MEDICINE & HEALTH SCIENCES, 15 March 2018 (2018-03-15), pages 1 - 77, XP055746083, ISSN: 1674-0246 * |
Also Published As
| Publication number | Publication date |
|---|---|
| US11816180B2 (en) | 2023-11-14 |
| US20220180095A1 (en) | 2022-06-09 |
| CN110059757A (zh) | 2019-07-26 |
| CN110059757B (zh) | 2021-04-09 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11972213B2 (en) | Event recognition method and apparatus, model training method and apparatus, and storage medium | |
| CN110798417B (zh) | 一种基于循环残差网络的信号调制识别方法及装置 | |
| US9769190B2 (en) | Methods and apparatus to identify malicious activity in a network | |
| US10984818B2 (en) | Devices and methods for evaluating speech quality | |
| US10547618B2 (en) | Method and apparatus for setting access privilege, server and storage medium | |
| CN111614398B (zh) | 基于异或神经网络的调制格式及信噪比识别方法及装置 | |
| US10827981B2 (en) | System and method for evaluating a cognitive load on a user corresponding to a stimulus | |
| US10691585B2 (en) | Efficient software testing system | |
| WO2021089013A1 (zh) | 空间图卷积网络的训练方法、电子设备及存储介质 | |
| WO2019227574A1 (zh) | 语音模型训练方法、语音识别方法、装置、设备及介质 | |
| CN107967488B (zh) | 一种服务器的分类方法及分类系统 | |
| CN111835763B (zh) | 一种dns隧道流量检测方法、装置及电子设备 | |
| CN111179971A (zh) | 无损音频检测方法、装置、电子设备及存储介质 | |
| EP3620982B1 (en) | Sample processing method and device | |
| WO2021082469A1 (zh) | 一种调制方式识别方法及装置 | |
| WO2021143126A1 (zh) | 一种调制方式确定方法及装置 | |
| WO2020215911A1 (zh) | 混合信号的分类方法、装置及电子设备 | |
| CN114155875B (zh) | 语音场景篡改鉴别的方法、装置、电子设备及存储介质 | |
| CN111581499B (zh) | 一种数据归一化方法、装置、设备及可读存储介质 | |
| US20220318689A1 (en) | Systems and Methods for Model Selection | |
| CN113921030B (zh) | 一种基于加权语音损失的语音增强神经网络训练方法及装置 | |
| TWI775186B (zh) | 射頻指紋訊號處理裝置及射頻指紋訊號處理方法 | |
| CN115049003A (zh) | 一种预训练模型微调方法、装置、设备和存储介质 | |
| CN111814051B (zh) | 一种资源类型确定方法及装置 | |
| CN112217749A (zh) | 一种盲信号分离方法及装置 |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| 121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 20796272 Country of ref document: EP Kind code of ref document: A1 |
|
| NENP | Non-entry into the national phase |
Ref country code: DE |
|
| 122 | Ep: pct application non-entry in european phase |
Ref document number: 20796272 Country of ref document: EP Kind code of ref document: A1 |