CN112380928A - Modulated signal classification method and system based on round system limited traversing visual graph networking - Google Patents
Modulated signal classification method and system based on round system limited traversing visual graph networking Download PDFInfo
- Publication number
- CN112380928A CN112380928A CN202011191624.8A CN202011191624A CN112380928A CN 112380928 A CN112380928 A CN 112380928A CN 202011191624 A CN202011191624 A CN 202011191624A CN 112380928 A CN112380928 A CN 112380928A
- Authority
- CN
- China
- Prior art keywords
- channel
- modulation signal
- classification
- modulation
- signal
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 58
- 230000000007 visual effect Effects 0.000 title claims abstract description 42
- 230000006855 networking Effects 0.000 title claims abstract description 26
- 239000013598 vector Substances 0.000 claims abstract description 40
- 238000013145 classification model Methods 0.000 claims abstract description 28
- 238000000605 extraction Methods 0.000 claims abstract description 23
- 238000012545 processing Methods 0.000 claims abstract description 20
- 230000004927 fusion Effects 0.000 claims abstract description 18
- 238000012549 training Methods 0.000 claims abstract description 6
- 238000010586 diagram Methods 0.000 claims description 19
- 238000010276 construction Methods 0.000 claims description 15
- 238000007637 random forest analysis Methods 0.000 claims description 6
- 230000008569 process Effects 0.000 claims description 4
- 238000004422 calculation algorithm Methods 0.000 description 5
- 230000006854 communication Effects 0.000 description 3
- 238000004891 communication Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 239000000284 extract Substances 0.000 description 2
- 238000013507 mapping Methods 0.000 description 2
- 239000011159 matrix material Substances 0.000 description 2
- 238000007418 data mining Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 238000001914 filtration Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
Images
Classifications
-
- 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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/243—Classification techniques relating to the number of classes
- G06F18/24323—Tree-organised classifiers
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/25—Fusion techniques
- G06F18/253—Fusion techniques of extracted features
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/02—Preprocessing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/08—Feature extraction
Abstract
The modulated signal classification method based on round system limited crossing visual graph networking comprises the following steps: s1, collecting I/Q modulation signals, processing the collected I/Q modulation signals, and converting the two-channel I/Q modulation signals into four-channel signals; s2, converting the four-channel signals into weighted directed network graphs respectively by adopting a round system limited traversing visible graph networking method; s3, performing feature extraction on the four weighted directed network graphs to obtain four feature vectors, and performing space expansion on the feature vectors to obtain a fusion feature vector of each I/Q modulation signal; s4, training the modulation signal classification model, adjusting the hyper-parameters in the round system limited crossing visual graph networking method when the classification precision is smaller than a preset threshold, repeating the steps S2-S3 until the classification precision is larger than or equal to the preset threshold to obtain the trained modulation signal classification model, and completing the classification of the I/Q modulation signal through the trained classification model. The invention can improve the classification precision of the I/Q modulation signals.
Description
Technical Field
The invention relates to the technical field of I/Q modulation signal classification, in particular to a modulation signal classification method and a modulation signal classification system based on round system limited-crossing visual graph networking.
Background
Time series are widely present in the real world, such as the internet, communications, biology, finance, and other various fields. In the field of communication, the problem of identification of modulation modes of I/Q modulation signals is receiving wide attention, and how to extract hidden information from a large amount of signal data to improve classification accuracy is one of the basic tasks of data mining of the current modulation signals.
Aiming at the problem of modulation signal classification, in the prior art, I/Q modulation signals with lower signal-to-noise ratio, such as signal modulation modes below-10 dB, cannot be identified. In order to solve the technical problem, researchers couple the time series with the field of complex networks, and excavate the structural features of the time series by researching the topological structure of the network, so as to achieve the purpose of signal classification. There are many methods for analyzing time series by visual graph algorithm in the academic world, such as visual graph VG, horizontal visual graph HVG, and limited traversing visual graph LPVG, and experiments prove that these methods for establishing network graph from time series can extract and retain some basic features of time series. However, the network graphs obtained by these networking algorithms are relatively single, and it is impossible to build a network graph containing more effective information from the time series according to different tasks or user requirements. Meanwhile, there is no method in the prior art for applying a visual pattern algorithm to classification of radio modulation signals.
Therefore, it is desirable to provide a system and method for classifying I/Q modulated signals with high flexibility and accuracy.
Disclosure of Invention
The invention aims to provide a modulation signal classification method and a modulation signal classification system based on round-system limited-traversal visual graph networking, which are used for solving the technical problems in the prior art and can effectively improve the flexibility and the accuracy of I/Q modulation signal classification.
In order to achieve the purpose, the invention provides the following scheme: the modulated signal classification method based on circle finite traversal visual graph networking comprises the following steps:
s1, collecting I/Q modulation signals, processing the collected I/Q modulation signals, and converting the two-channel I/Q modulation signals into four-channel signals;
s2, converting the four-channel signals into weighted directed network graphs respectively by adopting a round system limited traversing visible graph networking method;
s3, performing feature extraction on the four weighted directed network graphs to obtain four feature vectors, and performing spatial expansion on the feature vectors to obtain a fusion feature vector of each I/Q modulation signal;
s4, constructing a modulation signal classification model based on random forests, training the modulation signal classification model by adopting the fusion feature vector, adjusting the hyper-parameters in the round system finite-crossing visual image networking method when the classification precision is less than a preset threshold value, repeating the steps S2-S3 until the classification precision is greater than or equal to the preset threshold value to obtain a trained modulation signal classification model, and completing the classification of the I/Q modulation signals through the trained classification model.
Preferably, the specific method for converting the I/Q modulation signal into a four-channel signal in step S1 is as follows:
and processing the time sequence of the I channel and the Q channel of each I/Q modulation signal to obtain amplitude data A and phase data W of each I/Q modulation signal, thereby obtaining the time sequence of four channels of each I/Q modulation signal, wherein the four channels are respectively the I channel, the Q channel, the A channel and the W channel.
Further, the method for acquiring the amplitude data A and the phase data W comprises the following steps: the amplitude data a is calculated as shown in equation (1):
in the formula, AiRepresenting amplitude data at the ith time point in channel A, IiRepresenting signal data at the ith time point in the I channel, QiSignal data representing an ith time point in the Q channel;
the phase data W is calculated by taking an I channel as a horizontal coordinate and a Q channel as a vertical coordinate, and the formula (2) is as follows:
in the formula, WiIndicating the phase data of the ith time point in the W channel, IiRepresenting signal data at the ith time point in the I channel, QiSignal data representing the ith time point in the Q channel.
Preferably, in step S2, the nodes of the weighted directed network graph are mapped by the time points of the time series, the directed continuous edge of the weighted directed network graph is obtained by adding one to the number of times that the circular system formed by the two nodes can be cut off, where the node with the larger signal value points to the node with the smaller signal value.
Preferably, the specific method for performing spatial expansion on the feature vector in step S3 includes:
performing feature space expansion on four K-dimensional feature vectors obtained by each I/Q modulation signal in a transverse splicing mode
The invention also provides a system of the modulation signal classification method based on round system limited crossing visual graph networking, which comprises a data acquisition and processing unit, a network graph construction unit, a feature extraction unit and a modulation signal classification unit; the data acquisition and processing unit, the network diagram construction unit, the feature extraction unit and the modulation signal classification unit are sequentially connected and feed data in a single direction, the modulation signal classification unit feeds a super-parameter adjustment control signal to the network diagram construction unit, and the network diagram construction unit, the feature extraction unit and the modulation signal classification unit form a cycle;
the data acquisition and processing unit acquires an I/Q modulation signal, processes the acquired I/Q modulation signal, converts the dual-channel I/Q modulation signal into a four-channel signal, and transmits the four-channel signal to the network diagram construction unit;
the network graph constructing unit adopts a round finite traversing visible graph network constructing method to respectively convert the four-channel signals into weighted directed network graphs and transmits the four weighted directed network graphs to the feature extracting unit; the network graph constructing unit is also used for adjusting the hyper-parameters of the round system limited traversing visual graph network constructing method according to the control signal;
the feature extraction unit is used for extracting features of the four weighted directed network graphs to obtain four feature vectors, performing spatial expansion on the feature vectors to obtain a fusion feature vector of each I/Q modulation signal, and transmitting the fusion feature vector of each I/Q modulation signal to the modulation signal classification unit;
the modulation signal classification unit builds a modulation signal classification model based on random forests, trains the modulation signal classification model by adopting the fusion characteristic vector, sends a super-parameter adjustment control signal to the network diagram building unit until the classification precision is greater than or equal to a preset threshold value to obtain the trained modulation signal classification model, and finishes classification of the I/Q modulation signals through the trained modulation signal classification model.
The invention discloses the following technical effects:
the I/Q modulation signal time sequence classification task is converted into the network map classification task by adopting a round system limited crossing visual map networking method, in the training process of a classification model, the network map containing more effective information is obtained from the I/Q modulation signal time sequence by adjusting the hyper-parameter setting in the round system limited crossing visual map networking method, the potential structure information of the network map can be fully utilized, the classification precision of the I/Q modulation signal is improved, the modulation type of the signal can be accurately classified from the I/Q modulation signal, and a technical basis is provided for the subsequent processing of the signal and the information in the acquired signal; meanwhile, the method can be suitable for different types of I/Q modulation signals by adjusting the setting of the super parameters, and the flexibility is greatly improved.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings needed to be used in the embodiments will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings without inventive exercise.
FIG. 1 is a schematic diagram of a modulated signal classification system for round-based finite-pass visual graph networking according to the present invention;
FIG. 2 is a flowchart of a modulated signal classification method for round-based finite-traversal visual graph networking according to the present invention;
FIG. 3 is a schematic diagram of circle limited traversal visual graph networking in an embodiment of the present invention;
FIG. 4 is a schematic view of an arc formed by two points and a chord line connecting the two points according to an embodiment of the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
In order to make the aforementioned objects, features and advantages of the present invention comprehensible, embodiments accompanied with figures are described in further detail below.
Referring to fig. 1, the present embodiment provides a modulated signal classification system based on round-robin limited-traversal visual graph networking, for classifying I/Q modulated signals in a communication process, including: the system comprises a data acquisition and processing unit, a network diagram construction unit, a feature extraction unit and a modulation signal classification unit; the data acquisition and processing unit, the network diagram construction unit, the feature extraction unit and the modulation signal classification unit are sequentially connected and feed data in a single direction, the modulation signal classification unit feeds a super-parameter adjustment control signal to the network diagram construction unit in a connecting way, and the network diagram construction unit, the feature extraction unit and the modulation signal classification unit form a cycle;
the data acquisition and processing unit is used for acquiring I/Q modulation signals, processing the acquired I/Q modulation signals, converting the two-channel I/Q modulation signals into four-channel signals and transmitting the four-channel signals to the network diagram construction unit;
in this embodiment, the I/Q modulation signals in the training samples are obtained from a RADIOML 2016.10a data set, the RADIOML 2016.10a data set is a synthetic data set generated by GNU Radio, and the data set is first released at a sixth GNU year Radio conference; in the actual classification process, the data acquisition and processing unit receives an air radio signal through a receiving antenna of a radio receiver, and then an I/Q modulation signal is obtained through filtering and IQ demodulation.
The I/Q modulation signal is a two-channel time sequence comprising an I channel and a Q channel; the specific method for converting the I/Q modulation signal into a four-channel signal is as follows:
and processing the time sequence of the I channel and the Q channel of each I/Q modulation signal to obtain amplitude data A and phase data W of each I/Q modulation signal, thereby obtaining the time sequence of four channels of each I/Q modulation signal, wherein the four channels are respectively the I channel, the Q channel, the A channel and the W channel.
Wherein the amplitude data a is calculated as shown in equation (1):
in the formula, AiRepresenting amplitude data at the ith time point in channel A, IiRepresenting signal data at the ith time point in the I channel, QiSignal data representing an ith time point in the Q channel;
the phase data W is calculated by taking an I channel as a horizontal coordinate and a Q channel as a vertical coordinate, and the formula (2) is as follows:
in the formula, WiIndicating the phase data of the ith time point in the W channel, IiRepresenting signal data at the ith time point in the I channel, QiSignal data representing the ith time point in the Q channel.
The network graph constructing unit adopts a round finite traversing visible graph network constructing method to respectively convert the four-channel signals into weighted directed network graphs and transmits the four weighted directed network graphs to the feature extracting unit; the network graph constructing unit can also adjust the hyper-parameters of the round system limited traversing visual graph network constructing method according to the control signal;
in this embodiment, the network graph constructing unit converts the time series of the I channel, the Q channel, the a channel, and the W channel into the weighted directed network graph G by using a round-system finite-traversal visual graph constructing methodI、GQ、GAAnd GWThe nodes of the weighted directed network graph are formed by mapping the time points of the time sequence, the directed connecting edge of the weighted directed network graph is obtained by adding one to the number of times that the round system sight line formed by the two nodes can be cut off, wherein the node with the larger signal value points to the node with the smaller signal value in the two nodes, and the weight of the connecting edge of the weighted directed network graph can be effectively prevented from being 0 by adding one. The specific method for respectively converting the four-channel signals into the weighted directed network graph comprises the following steps:
converting the time series Y to { Y ═ Yi}i＝1,2,...,NSelecting points separated by l data points as nodes of the weighted directed network graph, wherein N represents the number of the data points in the time sequence Y; two nodes (t)m,ym) And (t)n,yn) The limited crossing sight distance between the two points is L, the actual crossing sight distance is M, wherein L represents the maximum allowable truncation times between the two points, M represents the actual truncation times between the two points, tmRepresents the data point ymCorresponding point in time, tnRepresents the data point ynA corresponding point in time; if the two nodes meet the visibility criterion, the connecting line of the two nodes forms the connecting edge of the weighted directed network graph, and the weight of the connecting edge is the actual crossing sight distance M +1, as shown in fig. 3.
Wherein the visibility criterion is described as:
as shown in fig. 4, with a node (t)m,ym) And (t)n,yn) The connecting line segment of (2) is a chord and forms an arc as a visual line, and the formula (3) is as follows:
f(t,y')＝(t-tm)(t-tn)+(y'-ym)(y'-yn)+a[(t-tm)(yn-ym)-(y'-ym)(tn-tm)]＝0…(3)
in the formula, t represents time, y' represents the value of a data point on a visual line, alpha represents a hyper-parameter, the size of alpha is adjustable, when alpha is more than 0, the visual line is concave, and when alpha is less than 0, the visual line is convex. In fig. 4, dotted lines indicate more edges than the view VG, and thick dashed lines indicate more edges than the LPVG.
By adjusting the size of alpha, the arc visual lines with different radians can be obtained; calculating the time t according to equation (3)k(m < k < n) is the value y 'on the visible line'k(ii) a Due to y'kAt time tkTwo solutions exist, and the solution on the circular arc less than 180 degrees is selected to be marked as y 'for ensuring visual significance'k；
If node (t)m,ym) And (t)n,yn) There are K data points in between (t)k,yk) K is not less than 0 and not more than L, satisfies yk>y'kThe remaining l-K data points (t)k,yk) Satisfy yk＜y'kI.e. at most no more than L time points are truncated, leaving L-K time points that are not allowed to be truncated, the visibility criterion is fulfilled.
The characteristic extraction unit is used for extracting the characteristics of the four weighted directed network graphs to obtain four characteristic vectors, performing spatial expansion on the characteristic vectors to obtain a fusion characteristic vector of each I/Q modulation signal, and transmitting the fusion characteristic vector of each I/Q modulation signal to the modulation signal classification unit;
in this embodiment, the feature extraction unit respectively weights four weights of each I/Q modulation signal by using a feature extraction methodDirected network graph GI、GQ、GAAnd GWPerforming feature extraction to obtain four K-dimensional feature vectors
And
wherein R isKA matrix representing a K dimension; performing feature space expansion on four K-dimensional feature vectors obtained by each I/Q modulation signal in a transverse splicing mode to obtain a fusion feature vector of each I/Q modulation signal
Where merge () represents merge.
In the embodiment, the feature extraction method adopts a Graph2vec algorithm in network embedding; graph2vec is an unsupervised embedding method of the first whole network, and the method is based on expanded characters and an embedding technology and shows great advantages in NLP; establishing a relation between the network and the rooted child Graph by using a model similar to doc2vec by the Graph2 vec; graph2vec first extracts rooted subgraphs and provides corresponding labels to put in a vocabulary, and then trains the Skip-Gram model to obtain a representation of the entire network. The method is excellent in network graph classification tasks and is superior to the classification effect of most artificial features in a plurality of graph classification tasks.
The modulation signal classification unit builds a modulation signal classification model based on random forests, trains the modulation signal classification model by adopting the fusion characteristic vector, sends a super-parameter adjustment control signal to the network diagram building unit until the classification precision is greater than or equal to a preset threshold value to obtain the trained modulation signal classification model, and finishes classification of the I/Q modulation signals through the trained modulation signal classification model.
Referring to fig. 2, the present embodiment further provides a modulated signal classification method based on round-system finite-crossing visual graph networking, which specifically includes the following steps:
s1, collecting I/Q modulation signals, processing the collected I/Q modulation signals, and converting the two-channel I/Q modulation signals into four-channel signals;
the I/Q modulation signal is a two-channel time sequence comprising an I channel and a Q channel; the specific method for converting the two-channel I/Q modulation signal into the four-channel signal comprises the following steps:
and processing the time sequence of the I channel and the Q channel of each I/Q modulation signal to obtain amplitude data A and phase data W of each I/Q modulation signal, thereby obtaining the time sequence of four channels of each I/Q modulation signal, wherein the four channels are respectively the I channel, the Q channel, the A channel and the W channel.
Wherein the amplitude data a is calculated as shown in equation (1):
in the formula, AiRepresenting amplitude data at the ith time point in channel A, IiRepresenting signal data at the ith time point in the I channel, QiSignal data representing an ith time point in the Q channel;
the phase data W is calculated by taking an I channel as a horizontal coordinate and a Q channel as a vertical coordinate, and the formula (2) is as follows:
in the formula, WiIndicating the phase data of the ith time point in the W channel, IiRepresenting signal data at the ith time point in the I channel, QiSignal data representing the ith time point in the Q channel.
S2, converting the four-channel signals into weighted directed network graphs respectively by adopting a round system limited traversing visible graph networking method;
the nodes of the weighted directed network graph are formed by mapping time points of the time sequence, the directed connecting edge of the weighted directed network graph is obtained by adding one to the number of times that the round system sight line formed by the two nodes can be cut off, wherein the node with the larger signal value points to the node with the smaller signal value.
The specific method for respectively converting the four-channel signals into the weighted directed network graph comprises the following steps:
converting the time series Y to { Y ═ Yi}i＝1,2,...,NSelecting points separated by l data points as nodes of the weighted directed network graph, wherein N represents the number of the data points in the time sequence Y; two nodes (t)m,ym) And (t)n,yn) The limited crossing sight distance between the two points is L, the actual crossing sight distance is M, wherein L represents the maximum allowable truncation times between the two points, M represents the actual truncation times between the two points, tmRepresents the data point ymCorresponding point in time, tnRepresents the data point ynA corresponding point in time; and if the two nodes meet the visibility criterion, the connecting line of the two nodes forms the connecting edge of the weighted directed network graph, and the weight of the connecting edge is the actual crossing sight distance M + 1.
Wherein the visibility criterion is described as:
with node (t)m,ym) And (t)n,yn) The connecting line segment of (2) is a chord and forms an arc as a visual line, and the formula (3) is as follows:
f(t,y')＝(t-tm)(t-tn)+(y'-ym)(y'-yn)+a[(t-tm)(yn-ym)-(y'-ym)(tn-tm)]＝0…(3)
in the formula, t represents time, y' represents the value of a data point on a visual line, alpha represents a hyper-parameter, the size of alpha is adjustable, when alpha is more than 0, the visual line is concave, and when alpha is less than 0, the visual line is convex.
By adjusting the size of alpha, the arc visual lines with different radians can be obtained; calculating the time t according to equation (3)k(m < k < n) is the value y 'on the visible line'k(ii) a Due to y'kAt time tkTwo solutions exist, in order to ensure the visual significance,selecting solutions on circular arcs smaller than 180 degrees as y'k，
If node (t)m,ym) And (t)n,yn) There are K data points in between (t)k,yk) K is not less than 0 and not more than L, satisfies yk>y'kThe remaining l-K data points (t)k,yk) Satisfy yk＜y'kThen the visibility criteria are met.
And S3, performing feature extraction on the four weighted directed network graphs to obtain four feature vectors, and performing spatial expansion on the feature vectors to obtain a fusion feature vector of each I/Q modulation signal.
In this embodiment, a feature extraction method is adopted to respectively apply four weighted directed network graphs G of each I/Q modulation signalI、GQ、GAAnd GWPerforming feature extraction to obtain four K-dimensional feature vectors
And
wherein R isKA matrix representing a K dimension; performing feature space expansion on four K-dimensional feature vectors obtained by each I/Q modulation signal in a transverse splicing mode to obtain a fusion feature vector of each I/Q modulation signal
Wherein merge () represents merge; in this embodiment, the feature extraction method uses a Graph2vec algorithm in network embedding.
S4, constructing a modulation signal classification model based on random forests, training the modulation signal classification model by adopting the fusion feature vector, adjusting the hyper-parameters in the round system finite-crossing visual image networking method when the classification precision is less than a preset threshold value, repeating the steps S2-S3 until the classification precision is greater than or equal to the preset threshold value to obtain a trained modulation signal classification model, and completing the classification of the I/Q modulation signals through the trained classification model.
The above-described embodiments are merely illustrative of the preferred embodiments of the present invention, and do not limit the scope of the present invention, and various modifications and improvements of the technical solutions of the present invention can be made by those skilled in the art without departing from the spirit of the present invention, and the technical solutions of the present invention are within the scope of the present invention defined by the claims.
Claims (6)
1. The modulated signal classification method based on circle finite traversal visual graph networking comprises the following steps:
s1, collecting I/Q modulation signals, processing the collected I/Q modulation signals, and converting the two-channel I/Q modulation signals into four-channel signals;
s2, converting the four-channel signals into weighted directed network graphs respectively by adopting a round system limited traversing visible graph networking method;
s3, performing feature extraction on the four weighted directed network graphs to obtain four feature vectors, and performing spatial expansion on the feature vectors to obtain a fusion feature vector of each I/Q modulation signal;
s4, constructing a modulation signal classification model based on random forests, training the modulation signal classification model by adopting the fusion feature vector, adjusting the hyper-parameters in the round system finite-crossing visual image networking method when the classification precision is less than a preset threshold value, repeating the steps S2-S3 until the classification precision is greater than or equal to the preset threshold value to obtain a trained modulation signal classification model, and completing the classification of the I/Q modulation signals through the trained classification model.
2. The method of claim 1, wherein the method comprises the steps of: the specific method for converting the I/Q modulation signal into a four-channel signal in step S1 is as follows:
and processing the time sequence of the I channel and the Q channel of each I/Q modulation signal to obtain amplitude data A and phase data W of each I/Q modulation signal, thereby obtaining the time sequence of four channels of each I/Q modulation signal, wherein the four channels are respectively the I channel, the Q channel, the A channel and the W channel.
3. The method of claim 2, wherein the circle-based finite traversal visual graph networking-based modulated signal classification method comprises: the method for acquiring the amplitude data A and the phase data W comprises the following steps: the amplitude data a is calculated as shown in equation (1):
in the formula, AiRepresenting amplitude data at the ith time point in channel A, IiRepresenting signal data at the ith time point in the I channel, QiSignal data representing an ith time point in the Q channel;
the phase data W is calculated by taking an I channel as a horizontal coordinate and a Q channel as a vertical coordinate, and the formula (2) is as follows:
in the formula, WiIndicating the phase data of the ith time point in the W channel, IiRepresenting signal data at the ith time point in the I channel, QiSignal data representing the ith time point in the Q channel.
4. The method of claim 1, wherein the method comprises the steps of: the nodes of the weighted directed network graph in step S2 are mapped by the time points of the time series, the directed connecting edge of the weighted directed network graph is obtained by adding one to the number of times that the round-system visible line formed by the two nodes is cut off, where the node with the larger signal value points to the node with the smaller signal value.
5. The method of claim 1, wherein the method comprises the steps of: the specific method for performing spatial expansion on the feature vector in step S3 is as follows:
and performing feature space expansion on four K-dimensional feature vectors obtained by each I/Q modulation signal in a transverse splicing mode.
6. The system for implementing the modulated signal classification method based on round-robin limited traversal visual graph networking of claim 1, wherein: the system comprises a data acquisition and processing unit, a network diagram construction unit, a feature extraction unit and a modulation signal classification unit; the data acquisition and processing unit, the network diagram construction unit, the feature extraction unit and the modulation signal classification unit are sequentially connected and feed data in a single direction, the modulation signal classification unit feeds a super-parameter adjustment control signal to the network diagram construction unit, and the network diagram construction unit, the feature extraction unit and the modulation signal classification unit form a cycle;
the data acquisition and processing unit acquires an I/Q modulation signal, processes the acquired I/Q modulation signal, converts the dual-channel I/Q modulation signal into a four-channel signal, and transmits the four-channel signal to the network diagram construction unit;
the network graph constructing unit adopts a round finite traversing visible graph network constructing method to respectively convert the four-channel signals into weighted directed network graphs and transmits the four weighted directed network graphs to the feature extracting unit; the network graph constructing unit is also used for adjusting the hyper-parameters of the round system limited traversing visual graph network constructing method according to the control signal;
the feature extraction unit is used for extracting features of the four weighted directed network graphs to obtain four feature vectors, performing spatial expansion on the feature vectors to obtain a fusion feature vector of each I/Q modulation signal, and transmitting the fusion feature vector of each I/Q modulation signal to the modulation signal classification unit;
the modulation signal classification unit builds a modulation signal classification model based on random forests, trains the modulation signal classification model by adopting the fusion characteristic vector, sends a super-parameter adjustment control signal to the network diagram building unit until the classification precision is greater than or equal to a preset threshold value to obtain the trained modulation signal classification model, and finishes classification of the I/Q modulation signals through the trained modulation signal classification model.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011191624.8A CN112380928B (en) | 2020-10-30 | 2020-10-30 | Modulation signal classification method and system based on circular finite-crossing visual networking |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011191624.8A CN112380928B (en) | 2020-10-30 | 2020-10-30 | Modulation signal classification method and system based on circular finite-crossing visual networking |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112380928A true CN112380928A (en) | 2021-02-19 |
| CN112380928B CN112380928B (en) | 2024-04-02 |
Family
ID=74576492
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011191624.8A Active CN112380928B (en) | 2020-10-30 | 2020-10-30 | Modulation signal classification method and system based on circular finite-crossing visual networking |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112380928B (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113392731A (en) * | 2021-05-31 | 2021-09-14 | 浙江工业大学 | Modulated signal classification method and system based on graph neural network |
| CN116524723A (en) * | 2023-06-27 | 2023-08-01 | 天津大学 | Truck track anomaly identification method and system |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20180357542A1 (en) * | 2018-06-08 | 2018-12-13 | University Of Electronic Science And Technology Of China | 1D-CNN-Based Distributed Optical Fiber Sensing Signal Feature Learning and Classification Method |
| CN110584596A (en) * | 2019-07-15 | 2019-12-20 | 天津大学 | Sleep stage classification method based on dual-input convolutional neural network and application |
| CN111707995A (en) * | 2020-06-19 | 2020-09-25 | 雷震烁 | Radar antenna scanning mode identification method based on limited-penetration visual effect |
-
2020
- 2020-10-30 CN CN202011191624.8A patent/CN112380928B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20180357542A1 (en) * | 2018-06-08 | 2018-12-13 | University Of Electronic Science And Technology Of China | 1D-CNN-Based Distributed Optical Fiber Sensing Signal Feature Learning and Classification Method |
| CN110584596A (en) * | 2019-07-15 | 2019-12-20 | 天津大学 | Sleep stage classification method based on dual-input convolutional neural network and application |
| CN111707995A (en) * | 2020-06-19 | 2020-09-25 | 雷震烁 | Radar antenna scanning mode identification method based on limited-penetration visual effect |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113392731A (en) * | 2021-05-31 | 2021-09-14 | 浙江工业大学 | Modulated signal classification method and system based on graph neural network |
| CN116524723A (en) * | 2023-06-27 | 2023-08-01 | 天津大学 | Truck track anomaly identification method and system |
| CN116524723B (en) * | 2023-06-27 | 2023-09-12 | 天津大学 | Truck track anomaly identification method and system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112380928B (en) | 2024-04-02 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111797716B (en) | Single target tracking method based on Siamese network | |
| CN110837778B (en) | Traffic police command gesture recognition method based on skeleton joint point sequence | |
| CN110298266A (en) | Deep neural network object detection method based on multiple dimensioned receptive field Fusion Features | |
| CN111832417B (en) | Signal modulation pattern recognition method based on CNN-LSTM model and transfer learning | |
| CN107292813A (en) | A kind of multi-pose Face generation method based on generation confrontation network | |
| CN102256065B (en) | Automatic video condensing method based on video monitoring network | |
| CN112380928A (en) | Modulated signal classification method and system based on round system limited traversing visual graph networking | |
| CN106683091A (en) | Target classification and attitude detection method based on depth convolution neural network | |
| CN106023177A (en) | Thunderstorm cloud cluster identification method and system for meteorological satellite cloud picture | |
| CN105307264B (en) | A kind of mobile node positioning method of wireless sensor network | |
| CN105374033A (en) | SAR image segmentation method based on ridgelet deconvolution network and sparse classification | |
| CN109508585A (en) | A method of urban function region is extracted based on POI and high-resolution remote sensing image | |
| CN105427313B (en) | SAR image segmentation method based on deconvolution network and adaptive inference network | |
| CN107392131A (en) | A kind of action identification method based on skeleton nodal distance | |
| CN112380931B (en) | Modulation signal classification method and system based on sub-graph network | |
| CN110149207A (en) | New type of continuous variable quantum key delivering method based on machine learning | |
| CN110298914A (en) | A kind of method of fruit tree canopy characteristic map in orchard establishing | |
| CN113392731B (en) | Modulation signal classification method and system based on graph neural network | |
| CN109117717A (en) | A kind of city pedestrian detection method | |
| Zhang et al. | A dynamic hand gesture recognition algorithm based on CSI and YOLOv3 | |
| CN103218829B (en) | A kind of foreground extracting method being adapted to dynamic background | |
| Kim et al. | Formulating human mobility model in a form of continuous time Markov chain | |
| CN106357461A (en) | Measuring method for air traffic display complexity | |
| Ma et al. | Deep learning based cognitive radio modulation parameter estimation | |
| CN113596757A (en) | Rapid high-precision indoor fingerprint positioning method based on integrated width learning |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |